WO2023249903A1 - Abrasive systems and methods of use - Google Patents

Abstract

A robotic abrading system is presented that includes a robotic arm that causes a workpiece to contact an abrasive article according to a set of operational parameters for an abrasive operation. The system also includes a pre-operation topography capturing device that captures a pre-operation surface topography for a portion of the workpiece. The system also includes a post-operation topography capturing device that captures a post-operation surface topography for the portion of the workpiece. The system also includes a wear calculator that calculates a wear level of the abrasive article after the abrasive operation. The system also includes a controller that generates an abrasive strategy based on the pre-operation surface topography, the abrasive strategy comprising the set of operational parameters, and the calculated wear level.

Description

ABRASIVE SYSTEMS AND METHODS OE USE

BACKGROUND

Abrasive articles are useful for shaping, finishing, or grinding a wide variety of materials and surfaces such as wood, metals (e g., especially non-ferrous metals such as aluminum that tend to clog grinding wheels), and flash. There continues to be a need for improving the cost, performance, and/or life of coated abrasive articles.

SUMMARY

A robotic abrading system is presented that includes a robotic arm that causes a workpiece to contact an abrasive article according to a set of operational parameters for an abrasive operation. The system also includes a pre-operation topography capturing device that captures a pre-operation surface topography for a portion of the workpiece. The system also includes a post-operation topography capturing device that captures a post-operation surface topography for the portion of the workpiece. The system also includes a wear calculator that calculates a wear level of the abrasive article after the abrasive operation. The system also includes a controller that generates an abrasive strategy based on the preoperation surface topography, the abrasive strategy comprising the set of operational parameters, and the calculated wear level.

Described herein are systems and methods for detecting when an abrasive article is nearing the end of its useful life. Some systems and methods herein may improve efficiency of use of abrasive articles as abrasive particles are worn down. However, systems and methods herein are not limited to measuring wear of abrasive particles. It may also be useful to detect wear of a resin matrix, for example in a nonwoven or bonded abrasive article where the entire article wears down during use. Some systems and methods herein may be particularly useful for robotic abrading systems, where a human operator is not available to detect the end of life by noticing the change in abrading efficiency.

The above summary of the present disclosure is not intended to describe each disclosed embodiment or every implementation of the present disclosure. The description that follows more particularly exemplifies illustrative embodiments. It is to be understood, therefore, that the following description should not be read in a manner that would unduly limit the scope of this disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a top-down schematic of an exemplary coated abrasive article.

FIG. 2 is a schematic cross-sectional view of an exemplary coated abrasive article.

FIG. 3 illustrates a robotic abrading system that may benefit from embodiments herein.

FIGS. 4A-4C illustrates a turbine blade and abrading strategy.

FIG. 5 illustrates a method of evaluating abrasive article efficacy in accordance with embodiments herein.

FIG. 6 illustrates a closed loop feedback system for a robotic abrading system in accordance with embodiments herein.

FIGS. 7 and 8A-8C illustrate a robot-centric cell for abrading in accordance with embodiments herein.

FIGS. 9A-9B illustrate bristle brush abrasive operations that may benefit from embodiments herein.

FIGS. 10A-10B illustrate an abrasive operation using an abrasive wheel that may benefit from embodiments herein.

FIG. 11 illustrates a method of abrading a component using an automated abrading system in accordance with embodiments herein.

FIGS. 12A-12B illustrate methods of evaluating a wear level of an abrasive article in accordance with embodiments herein.

FIG. 13 illustrates a robotic abrading system in accordance with embodiments herein.

FIG. 14 is a defect inspection system architecture.

FIG. 15 illustrates an example computing device that can be used in embodiments shown in previous Figures.

FIGS. 16 and 17A-17D illustrate Examples as described in greater detail herein.

Repeated use of reference characters in the specification and drawings is intended to represent the same or analogous features or elements of the disclosure. It should be understood that numerous other modifications and embodiments can be devised by those skilled in the art, which fall within the scope and spirit of the principles of the disclosure. The figures may not be drawn to scale. DETAILED DESCRIPTION

Throughout this document, values expressed in a range format should be interpreted in a flexible manner to include not only the numerical values explicitly recited as the limits of the range, but also to include all the individual numerical values or sub-ranges encompassed within that range as if each numerical value and sub-range is explicitly recited. For example, a range of “about 0.1% to about 5%” or “about 0.1% to 5%” should be interpreted to include not just about 0.1% to about 5%, but also the individual values (e.g., 1%, 2%, 3%, and 4%) and the sub-ranges (e.g., 0.1% to 0.5%, 1.1% to 2.2%, 3.3% to 4.4%) within the indicated range. The statement “about X to Y” has the same meaning as “about X to about Y,” unless indicated otherwise. Likewise, the statement “about X, Y, or about Z” has the same meaning as “about X, about Y, or about Z,” unless indicated otherwise.

In this document, the terms “a,” “an,” or “the” are used to include one or more than one unless the context clearly dictates otherwise. The term “or” is used to refer to a nonexclusive “or” unless otherwise indicated. The statement “at least one of A and B” has the same meaning as “A, B, or A and B.” In addition, it is to be understood that the phraseology or terminology employed herein, and not otherwise defined, is for the purpose of description only and not of limitation. Any use of section headings is intended to aid reading of the document and is not to be interpreted as limiting; information that is relevant to a section heading may occur within or outside of that particular section.

In the methods described herein, the acts can be carried out in any order without departing from the principles of the disclosure, except when a temporal or operational sequence is explicitly recited. Furthermore, specified acts can be earned out concurrently unless explicit claim language recites that they be carried out separately. For example, a claimed act of doing X and a claimed act of doing Y can be conducted simultaneously within a single operation, and the resulting process will fall within the literal scope of the claimed process.

The term “about” as used herein can allow for a degree of variability in a value or range, for example, within 10%, within 5%, or within 1% of a stated value or of a stated limit of a range, and includes the exact stated value or range.

The term “substantially” as used herein refers to a majority of, or mostly, as in at least about 50%, 60%, 70%, 80%, 90%, 95%, 96%, 97%, 98%, 99%, 99.5%, 99.9%, 99.99%, or at least about 99.999% or more, or 100%. As used herein, the term "shaped abrasive particle," means an abrasive particle with at least a portion of the abrasive particle having a predetermined shape that is replicated from a mold cavity used to form the shaped precursor abrasive particle. Except in the case of abrasive shards (e.g. as described in US Patent Application Publication Nos. 2009/0169816 and 2009/0165394), the shaped abrasive particle will generally have a predetermined geometric shape that substantially replicates the mold cavity that was used to form the shaped abrasive particle. Shaped abrasive particle as used herein excludes abrasive particles obtained by a mechanical crushing operation. Suitable examples for geometric shapes having at least one vertex include polygons (including equilateral, equiangular, star-shaped, regular and irregular polygons), lens- shapes, lune-shapes, circular shapes, semicircular shapes, oval shapes, circular sectors, circular segments, drop-shapes and hypocycloids (for example super elliptical shapes).

For the purposes of this invention, geometric shapes are also intended to include regular or irregular polygons or stars wherein one or more edges (parts of the perimeter of the face) can be arcuate (either of towards the inside or towards the outside, with the first alternative being preferred). Hence, for the purposes of this invention, triangular shapes also include three- sided polygons wherein one or more of the edges (parts of the perimeter of the face) can be arcuate. The second side may include (and preferably is) a second face. The second face may have a perimeter of a second geometric shape.

For the purposes of this invention, shaped abrasive particles also include abrasive particles comprising faces with different shapes, for example on different faces of the abrasive particle. Some embodiments include shaped abrasive particles with different shaped opposing sides. The different shapes may include, for example, differences in surface area of two opposing sides, or different polygonal shapes of two opposing sides.

The shaped abrasive particles are typically selected to have an edge length in a range of from 0.001 mm to 26 mm, more typically 0.1 mm to 10 mm, and more typically 0.5 mm to 5 mm, although other lengths may also be used.

However, it is expressly contemplated that systems and methods herein may be useful for abrasive articles that do not contain precisely shaped grain, or contain a mix of precisely shaped and crushed grain.

Described in embodiments here are systems and methods that track wear of an abrasive article and adjust operating parameters in response to the detected level of wear. Described herein are systems and methods that utilize abrasive belts. However it is expressly contemplated that abrasive discs, bonded abrasive wheels, and other suitable abrasive articles may also be used with such systems.

Human operators cannot see very low levels (e.g. urn) of material removed across complex geometry. Human operators are also not as sensitive to changes in acoustics, temperatures, vibrations as robotic systems can be. Systems and methods are needed that can sense such changes and use them as process inputs in adaptive process operating windows.

FIGS. 1 and 2 show an exemplary coated abrasive disc 100 according to the present disclosure, wherein shaped abrasive particles 130 are secured at precise locations and Z-axis rotational orientations to a backing 110. In one embodiment, shaped abrasive particles 130 are triangular prism shaped particles that appear rectangular when viewed from above. While FIG. 1 illustrates a coated abrasive disc, it is expressly contemplated that systems and methods herein may also use abrasive belts, nonwoven pads, bristle brushes or abrasive grinding wheels. For example, FIGS. 9-11 illustrate embodiments concerning other abrasive articles.

Generally, a coated abrasive article 100 includes a plurality of abrasive particles embedded within a make coat that secures the particles to a backing. The backing may be formed from any known flexible coated abrasive backing, for example. Suitable materials for the backing include polymeric films, metal foils, woven fabrics, knitted fabrics, paper, nonwovens, foams, screens, laminates, combinations thereof, and treated versions thereof.

The abrasive particles 130 may be embedded within an abrasive layer, which can include multilayer construction having make 120 and size layers 140. Coated abrasive articles according to the present disclosure may include additional layers such as, for example, an optional supersize layer that is superimposed on the abrasive layer, or a backing antistatic treatment layer may also be included, if desired. Exemplary suitable binders can be prepared from thermally curable resins, radiation-curable resins, and combinations thereof.

Make layer 120 can be formed by coating a curable make layer precursor onto a major surface of backing 110. The make layer precursor may include, for example, glue, phenolic resin, ammoplast resin, urea-formaldehyde resin, melamine-formaldehyde resin, urethane resin, free-radically polymerizable polyfunctional (meth)acrylate (e g., aminoplast resin having pendant a,|B-unsaturated groups, acrylated urethane, acrylated epoxy, acrylated isocyanurate), epoxy resin (including bis-maleimide and fluorene-modified epoxy resins), isocyanurate resin, and mixtures thereof. Of these, phenolic resins are preferred.

Phenolic resins are generally formed by condensation of phenol and formaldehyde, and are usually categorized as resole or novolac phenolic resins. Novolac phenolic resins are acid-catalyzed and have a molar ratio of formaldehyde to phenol of less than 1 :1. Resole (also resol) phenolic resins can be catalyzed by alkaline catalysts, and the molar ratio of formaldehyde to phenol is greater than or equal to one, typically between 1.0 and 3.0, thus presenting pendant methylol groups. Alkaline catalysts suitable for catalyzing the reaction between aldehyde and phenolic components of resole phenolic resins include sodium hydroxide, barium hydroxide, potassium hydroxide, calcium hydroxide, organic amines, and sodium carbonate, all as solutions of the catalyst dissolved in water.

Resole phenolic resins are typically coated as a solution with water and/or organic solvent (e.g., alcohol). Typically, the solution includes about 70 percent to about 85 percent solids by weight, although other concentrations may be used. If the solids content is very low, then more energy is required to remove the water and/or solvent. If the solids content is very high, then the viscosity of the resulting phenolic resin is too high which typically leads to processing problems.

Phenolic resins are well-known and readily available from commercial sources. Examples of commercially available resole phenolic resins useful in practice of the present disclosure include those marketed by Durez Corporation under the trade designation VARCUM (e.g., 29217, 29306, 29318, 29338, 29353); those marketed by Ashland Chemical Co. of Bartow, Florida under the trade designation AEROFENE (e g., AEROFENE 295); and those marketed by Kangnam Chemical Company Ltd. of Seoul, South Korea under the trade designation PHENOLITE (e.g., PHENOLITE TD-2207).

The make layer precursor may be applied by any known coating method for applying a make layer to a backing such as, for example, including roll coating, extrusion die coating, curtain coating, knife coating, gravure coating, and spray coating.

The basis weight of the make layer utilized may depend, for example, on the intended use(s), type(s) of abrasive particles, and nature of the coated abrasive article being prepared, but typically will be in the range of from 1, 2, 5, 10, or 15 grams per square meter (gsm) to 20, 25, 100, 200, 300, 400, or even 600 gsm. The make layer may be applied by any known coating method for applying a make layer (e g., a make coat) to a backing, including, for example, roll coating, extrusion die coating, curtain coating, knife coating, gravure coating, and spray coating.

Once the make layer precursor is coated on the backing, the triangular abrasive particles are applied to and embedded in the make layer precursor. The triangular abrasive particles are applied nominally according to a predetermined pattern and Z-axis rotational orientation onto the make layer precursor. Using known orientation methods, such as electrostatic or magnetic orientation, it is possible to orient the abrasive particles with respect to the backing in order to improve performance of the particles.

However, while FIGS. 1-2 illustrate a coated abrasive article, it is expressly contemplated that systems and methods herein may also be suitable for understanding use and wear of other abrasive articles such as bonded abrasive articles with resin or vitreous bond matrices, nonwoven abrasive articles, brushes, or other abrasive articles.

Abrasive articles may be used in a number of contexts. Described herein is a particular configuration of a robotic abrading cell with a closed-circuit feedback system. However, it is expressly contemplated that systems and methods described herein can be implemented in different automated abrading configurations.

Different use scenarios of abrasive articles present different problems regarding article use over time. For example, an experienced human operator can often “feel” when an abrasive article is losing cut efficacy over time and adjust accordingly, by applying more force or adjusting an angle.

In contrast, a robotic system may have no insight into the wear or loading occurring on an abrasive article and may not make necessary adjustments, or replace an abrasive article when needed without intervention. For example, knowing an amount of wear on an abrasive article, process parameters can be modified to maintain abrasive efficacy throughout the service life of the abrasive article and / or while maintaining expected material removal rates. Systems and methods herein may be useful in other contexts as well.

FIG. 3 is a schematic of a robotic arm that may benefit from embodiments disclosed herein. A robotic repair unit 200 has a base 210, which may be stationary, in some embodiments. In other embodiments, base 210 can move in any of six dimensions, translations or rotations about an x-axis, y-axis and/or z-axis. For example, robot 200 may have a base 210 fixed to a rail system configured to travel along with a moving substrate being repaired. Depending on a particular operation, robot 200 may need to move closer, or further away from a substrate, or may need to move higher or lower with respect to an abrading area. A moveable base 200 may thus increase functionality.

Robotic arm unit 200 has one or more tools 240 that can interact with a worksurface. Tool 240 may include a backup pad 250, in one embodiment, or another suitable abrasive tool. During an abrasive operation, tool 240 may have an abrasive disc, or other suitable abrasive article, attached using adhesive, hook and loop, clip system, vacuum or other suitable attachment system. However, as the abrasive article moves in conjunction with a backup pad 250 to which it is attached, the abrasive article is not necessanly considered as adding additional degrees of freedom to the movement of robotic repair unit 200. As mounted to the robotic repair unit 200, tool 240 has the ability to be positioned within the provided degrees of freedom by the robotic repair unit 200 (6 degrees of freedom in most cases) and any other degrees of freedom (e g., a compliant force control 230 unit).

Backup pad 250 is coupled to a tool 240 which has an orbit that provides some additional degrees of freedom. In most tools a single degree of freedom is provided by a rotating shaft with or without some offset. Tool 240 is coupled to a force control 230 unit output. Force controlled flange 230 provides a soft (i.e., not stiff) displacement curve. In most force control units, a single degree of freedom is provided by a sliding (prismatic) joint along the active axis. Force control 230 is coupled to a flange 220. Movement of components 210, 220, 230, 240 and 250 is all controllable using a robot controller (e.g., robot controller 270).

Robotic controller 270, in addition to moving components 210-250 based on the parameters of an abrading operation, may also adjust parameters based on information received from an abrasive article use evaluation 260. For example, if evaluation system 260 indicates that an abrasive article has reached an end of life, controller 270 may instruct system 200 to stop an abrading operation, change out the old abrasive article for a new abrasive article, and then continue an abrading operation. Additionally, for example, controller 270 may provide new parameters for abrading based on the feedback from system 260. For example, if an abrasive article loaded, controller 270 may increase a coolant flow to flush accumulated swarf. If the abrasive article is capped, controller 270 may initiate a dressing process to alleviate detected metal capping. If the abrasive article is worn, but not at end of life, controller may increase a force applied by force control unit 230, or may adjust an angle of tool 240 with respect to a substrate. Such adjustments are described in greater detail with reference to later figures.

FIGS. 4A-4C illustrates a jet engine turbine blade and abrading strategy. FIG. 4A illustrates a molded part 400 with a number of gates 402 that need to be removed and a surface 404 that needs to be treated. Currently, jet engine turbines such as the finished product illustrated in FIG. 4C take over 14 hours of manual abrading, through multiple steps, to achieve a finished product. And, in many cases, over half of the finished products are discarded because they cannot be balanced properly. This is due, in part, to the fact that it is cheaper to finish the process by hand than to quality check in between steps. Metrologybased quality control is often contact-based which requires additional time. The process is time consuming, in part, because of the precision required for the final product to be balanced, and the different abrading processes that have to be done. Similar blades are also used for land-based power turbines. Other industries that may also benefit from systems and methods described herein may include other precision abrading operations, such as orthopedic implants. Other processes may also benefit from automation as described herein.

FIG. 4B illustrates an abrading strategy for the part 400. Any ceramic casting residue may be blasted off. Casting gates 412 are removed using grinding wheels and abrasive belts. The root gate is removed, which entails 3-4 mm of material removal. For gate removal, there- is a more lenient tolerance for material removal than in later steps.

In a second step 414, pin grinding smooths out the surface to +/- 40 pm from a final airfoil surface. It is important not to over machine at this point.

A final polishing step 416 is done before blasting, etching and tumbling. Polishing can be extended to the root and platform areas as well Previously this step involved hand dressing because the finish should be within 0.7 pm of specification. A consistent finish is desired over the airfoil surface. It may also be beneficial to extend the polished area to the root and platform areas. Better finishing during this stage saves time downstream in blasting, etching and tumbling steps.

The final result is illustrated in FIG. 4C. Turbine 430 has a precisely abraded, smooth surface. As noted previously, the current manual process can take over 14 hours and result in 50% of turbines 430 being scrapped at the end because of uncurable defects. A process with a higher yield rate is desired, and improved speed would also be preferred. Automation of precisely abraded parts, like turbine 430, has proved challenging, in part, because the wear rate of an abrasive article changes the material removal rate, and each time an abrasive article is used, abrasive particles fracture and are worn down. This can cause the same abrasive to perform differently during a 10^th use than a 1^st use. It may be possible to achieve the same performance in the 10^th use, for example, by adjusting parameters such as force or movement speed to compensate for the wear. For precise material removal on a contoured surface, it is important to have a real-time understanding of the current wear of an abrasive article.

It is also desired to obtain a wear indication for an abrasive article in-situ, during an abrading process. Offline systems, such as FEA, particle swarm or electron microscopy as this requires either an abrasive article or a part being abraded to be removed from the abrading operation, taken to a lab to be treated and imaged. Instead, an in-situ system is desired that can provide an indication of abrasive wear without interrupting an abrading operation and, preferably, without contacting the abrasive article. Described herein are systems and methods for providing in-situ feedback and real-time estimates of abrasive wear.

While the abrading process for turbine blades is described herein as one example process that can benefit from the sy stems and methods described herein, it is expressly contemplated that systems and methods herein can be used for other high-precision abrasive operations, like precision cast and forged components, plumbing fixtures or other components with multiple abrading steps that would benefit from intermediary step quantification.

FIG. 5 illustrates a method of evaluating abrasive article efficacy in accordance with embodiments herein. In some embodiments, evaluating may be automatically completed using method 500.

In block 510, an abrasive article is installed on a tool. For example, the abrasive article may be coupled to a backup pad or directly to a tool of a robotic abrading unit. The abrasive article may be any suitable abrasive article for a given abrading operation, such as a coated abrasive article, a bonded abrasive article, an abrasive belt, a bristle brush, or another suitable abrasive article. In block 520, the abrasive article engages a workpiece, and an abrading operation is conducted. As the workpiece is abraded, abrasive particles of the abrasive article experience wear. They may also experience metal capping, loading or other degradation.

In block 530, a post-abrasive evaluation is done. The post-abrasive evaluation may be done on the abrasive article 532, on a workpiece abraded 534, or in another manner 536. An abrading efficacy of the abrasive article is determined. Because the exact abrasive parameters are known, e.g. how the abrasive article contacted a workpiece, with what applied force, at what speed, it is possible to model how the abrasive article changes over time. However, in embodiments herein, the modeled wear is confirmed by examining the worksurface and comparing pre- and post-abrasion images to determine an actual amount of material removed. This can then be compared to the expected material removal rate and, based on the comparison, an actual wear can be calculated. Abrasive efficacy may also include an examination of the abrasive surface profile of the abrasive article and comparing it to a desired surface profile. For example, some abrasive articles are shaped in order to impart a particular shape to a workpiece.

In block 542, if the abrasive efficacy of the abrasive article is still at an acceptable level, then it can continue to be used for another abrasive operation. As noted with respect to FIG. 4B, some abrasive steps have tolerances for material removal rate, e.g. initial gate removal, etc. It may, therefore, acceptable to proceed with currently programmed parameters. In some embodiments, new operational parameters 544 are provided based on a detected decrease in abrasive efficacy between evaluations. The new parameters may be automatically implemented.

In block 546, if the abrasive efficacy is below an acceptable level, it can either be treated, as indicated in block 550, for example by dressing the abrasive article to remove metal capping, washing or cleaning the abrasive article to remove loading, or re-dressing to re-establish a surface profile desired. However, in some embodiments, method 500 returns to block 510 and a new abrasive article is installed.

FIG. 6 illustrates a closed loop feedback system for a robotic abrading system in accordance with embodiments herein. As described herein, a vision system captures pre and post abrasion inspection images in between steps of a multi-step or in between sequences of a multi-sequence operation. The information from the captured images can then correlate the appearance of either an abrasive article or a workpiece with an actual wear rate of the abrasive article.

Systems and methods herein collect data from multiple sources, and then structure the data for software analytics to predictively model abrasive wear rate and material removal rate. This will allow for both extended abrasive life through increased efficacy and accurate material removal. This will increase yield rates of correctly manufactured parts.

Thus, as illustrated in FIG. 6, a closed-circuit feedback system can improve abrasive efficacy and extend the life of abrasive articles. Robotic abrading system 610 includes one or more consumable abrasive articles 602 that contact and abrades a substrate 604. For example, as described with respect to FIG. 7, a robotic arm may bring a part 604 into contact with an abrasive belt 602. However, it is expressly contemplated that, in other embodiments, that a robotic arm brings an abrasive article into contact with a substrate.

System 610 may have sensors 606, which may include optical sensors, accelerometers, pressure sensors, as well as setting readers that can detect current settings of a tool. In some embodiments, sensors 610 include an optical scanning system that captures images pre and post abrading operation. For examples, sensors 610 may include one or more cameras or video cameras. System 610 may also have other features 608.

Abrading system 610 may provide information automatically to a datastore 630, for example sensor data 620 retrieved from sensors 606 of system 610 or from elsewhere in a production line (e g. a vision system upstream or downstream). Additionally, datastore 630 may receive settings data 620 from abrading system 610. Datastore 630 may also include information about an abrasive article used, parameter settings of system 610 during any or all steps of an abrasive operation, as well as feedback of an operator or other quality control system. Datastore 630 may also include material removal profiles 634, for example obtained by or generated using sensors 606 in a previous abrading operation. Abrasive wear profiles 632 may be simulated or constructed using historic data for a number of abrasive articles 602. For example, historic operational outcomes 636 may be stored in datastore 630. Operational outcomes 636 may include physical results of an abrasive operation and / or whether or not the final product was deemed acceptable. Substrate information 638 may also be stored, for example a material, an initial surface contour and / or a final surface contour. Datastore 630 may contain other information 639. Datastore 630 may also have other information, for example received from many different abrading systems 610 and / or from many different abrading operations.

An analyzer 650 may retrieve data 640 and job specifications 642 from datastore 630. Analyzer 650 may be part of a controller, e.g. such as an application controller, or may be communicably coupled to datastore 630 and / or abrading system 610. Job specifications may be retrieved based on an identification of a part to be abraded. Job specifications may be based on a desired final surface profile of the tool, or based on a detected defect that needs to be addressed. Analyzer 650 may have one or more machine learning algorithms 652 that generate parameter sets based on information in datastore 630. Analyzer 650 may be powered by a decision tree regressor algorithm, for example, which creates a decision tree that best represents the training data. However, other machine learning algorithms or techniques may be used.

While FIG. 6 illustrates a closed-circuit system 600 where information about abrading system 610 is stored in a datastore 630 and then retrieved from datastore 630 by analyzer 650, it is expressly contemplated that, in some embodiments, feedback loop 600 illustrates the steps performed by an algorithm stored on a processor of a controller. For example, as sensor data 620 is collected, an abrasive wear profile 632 is generated for an abrasive article 602 and, based on an amount of wear, settings are adjusted to maintain a desired matenal removal. For example, a dwell time, speed or force may be increased achieve a desired material removal.

Analyzer 650, based on retrieved information from datastore 630 and job specifications, may use machine learning techniques 652 to select parameter settings 664, within the limits of parameter ranges 658 to achieve the job specifications 642.

Controller 670 retrieves a set of settings 664 and generates a command 684 using command generator 672. Command 684, when received by robotic abrading system, changes operating parameters in-situ.

In some embodiments it may not be possible to immediately adjust settings. In some embodiments, command 684 includes a time delay or other indication of when settings should be changed, e.g. increase force gradually from a current setting to a new setting, or to wait until a next abrading step starts to increase the force.

FIGS. 7 and 8A-8C illustrate a robot-centric cell for abrading in accordance with embodiments herein. FIG. 7 illustrates a top-down view of a robot cell 700. Cell 700 includes all the components necessary for the processing of a turbine 400. However, it is expressly contemplated that, while robot 702 is illustrated in a cell with multiple abrading tools 722-726, it is expressly contemplated that systems and methods herein can be useful for a robot 702 that interacts with a single abrasive tool.

In robotic cell 700, a robot 702 is in a fixed position with a radial arm reach 750. However, it is expressly contemplated that robot 702, or any of the illustrated components, may be motive components. An operator places a turbine 700 on pickup location 704, where robot abrading system 702 picks it up. In a first abrading operation, the part is taken to a first abrading station 722, a spindle in the illustrated embodiment. Spindle 722 removes gates. Two back stands 724, 726 are illustrated, each having a different grade abrasive belt installed thereupon. In the illustrated embodiment, back stands 724, 726 and spindle 722 have a force control unit that modulates the force applied against a turbine held by robot 702. While three abrading units are illustrated in FIG. 7, it is expressly contemplated that only one abrasive article is present, two abrasive articles are present, as well as more than three, such as four or more. In a second abrading operation, robot 702 may take a part to abrading unit 724. In a third abrading operation, robot 702 may proceed to take the part to abrading unit 726.

While only a single belt is shown on abrading units 724, 726, it is expressly contemplated that each abrading unit may have multiple grades or types of abrading units such as a progression of coarse to fine grade belts, etc. The abrasive article may include microreplicated particles (e.g. sold under the brand name TRIZACT™) or diamond coated abrasive particles. Abrading units 724, 726 may have multiple belts and / or plattens.

Before and after some or all of abrasive operations, the component is imaged. Two imaging components 732, 734 are illustrated in FIG. 7. However, in other embodiments, only one imaging component, or more than two, are possible. In one embodiment, a first imaging device 732 is used for coarse removal of material while a second imaging device 734 is used for imaging after finer operations. In the illustrated embodiment of FIGS. 7-8, component 732 is a hexagon scanner and component 734 is a profilometer. However, it is expressly contemplated that other optical sensors may be used.

An initial scan is done before any abrading, by component 732, to understand the geometry of the component as it is being held by the robot. The position and relative configuration of the component is required to generate a trajectory for the robot - how to approach one of abrading devices 722, 724, or 726 so that the component is angled correctly. The initial scan may be done using a pre-set protocol. The scan is compared to a reference - e.g. a rendering of a “perfect” turbine or a desired final profile. Based on the scan, differences between a current part (e.g. part 400) and the desired final product are noted. In some embodiments, imaging component 732 has its own datastore and controller to manage the data captured during a scan and to analyze the scan to determine an abrading strategy to remove necessary material to obtain the final product.

In between each abrasive step, the component is brought back to either component 732 or 734 for an imaging step. This is done both to confirm that a previous abrasive operation was successful, and whether the component needs to “redo” a previous step to remove additional material removal, or whether the component can proceed to the next step. Based on what was, and was not successful, machine learning algorithms can be used to “learn” a better way to measure and react to abrasive article wear. Additionally, as abrasive article wear is better understood, it may be possible both to achieve better efficiency in each abrasive step, but also to reduce a wear rate and extend the service life of the abrasive article.

A second imaging component 734 may be used for finer operations. The second imaging component 734 may be used instead of, or in addition to, component 732. In some embodiments described herein, component 734 is a profilometer that identifies a topography. The two technologies used in the identification and characterization of the surface topography are confocal fusion and focus variation.

However, while a Hex scanner and an S Neox profilometer are illustrated, it is expressly contemplated that other imaging technology could be used, including, but not limited to 2D cameras, 3D cameras, video cameras, etc. In some embodiments, fringe projection, white light interferometry, deflectometry or structured light multi-stereo scanners may be used as well, or in addition to vision systems discussed herein.

A controller 760 may control robot 702 and one or more of components 722, 624, 726 and / or imaging components 732, 734. However, it is expressly contemplated that some or all of these components may have their own controllers.

Controller may apply machine learning algorithms to the scans obtained from imaging components 732, 734 to determine an order of abrading operations as well as parameter sets to use for and during each operation. For example, it may be suitable to remove gates first, or it may be beneficial to keep one gate attached until closer to the end so that the robot has a space to couple to (e g. “grab”) the component. One component may need 3 grinding steps to remove gates and large quantities of material while a different component may need 4.

FIGS. 8A-8B illustrate images of a robotic cell from a first view 800 and a second view 820. FIG. 8C illustrates a frontal view and a profile view of a scanning system 840 of a metrology system used for obtaining a topography of a component undergoing a sequence of abrasive operations.

FIGS. 9A-9B illustrate bristle brush abrasive operations that may benefit from embodiments herein. Some operations, such as gear formation, have specific final surface profiles. Using a bristle brush that is not shaped for a current worksurface curvature can lead to different contact and removal rates, and may result in a final worksurface profile that does not match the desired curvature. Due to centrifugal force from rotation of the brush while in use, the bristles extend outward from the brush, potentially resulting in an abrasive profile other than what might be anticipated for a robotic abrading system. Bristle brushes are intended to be flexible, but it can be important to ensure that contact profiles are as expected, especially for robotic abrading operations, where precise and accurate abrading profiles are required. Additionally, a bristle brush changes dimensions due to wear - bristles may break, or be worn down, which should be tracked in order to ensure that a desired wear rate and profile is maintained.

FIG. 9A illustrates an operation 900 that includes a bristle brush 902 that will contact and abrade worksurface 920. Brush 902 has a number of bristles 910, each containing abrasive particles or otherwise having an abrasive surface. When brush 902 is moved against surface 920, material is removed by abrading. However, bristles 910 will not all contact surface 920 at the same angle, due to the curved surface. This will cause uneven pressure, and uneven wear on bristles 910. Additionally, it may cause undesired and uneven wear on surface 920 as bristles are forced into unintended directions and some areas will experience more contact, and others less, than expected.

In FIG. 9B, in contrast, an operation 950 includes a bristle brush 952 with bristles 960 that have been shaped in advance such that bristle lengths are sized for a curve in surface 970.

FIGS. 10A-10B illustrate an abrasive operation using an abrasive wheel that may benefit from embodiments herein. Abrasive wheels can be formed from bonded abrasives - e g. abrasive particles in a resin, vitreous or polymeric bond - or using non woven material that contains abrasive particles The nonwoven material may be wrapped into a cylinder, which may then be cut into individual wheels. Such unitized abrasive wheels can be used in a number of applications.

Due to centrifugal force from rotation of the abrasive article during use, such wheels can extend in diameter. Additionally, there can be axial runout, which can lead to an unknown relative position of the abrasive article with respect to the robotic abrasive unit, which could cause an abrasive operation to be unsuccessful, off target, or cause damage to a workpiece. Therefore, it may be necessary or helpful to recondition abrasive wheels periodically. Removing a wheel from a robotic repair operation to recondition, cut or dress it, however, takes time and is not always feasible. Additionally, dimensional changes due to wear require adaptation of abrasive parameters to maintain a desired wear rate.

FIG. 10A illustrates an abrasive operation 1000 where an abrasive article 1010 is approaching contact with worksurface 1020. Article 1010 is sized to interact with a shaped surface of worksurface 1020. As abrasive article 1010 wears down, a z-position 1002 of the article with respect to surface 1020 changes. Additionally, a profile along the x-axis can change due to rotational velocity.

FIG. 10B illustrates a top-dow n view of an abrasive wheel 1050. Rotational velocity 1052 can change during an operation, causing expansion of the circumference of article 1050. Additionally, as the article is worn down, a surface condition 1052 can change.

In some embodiments, systems and methods herein include a dressing or treatment station that receives an abrasive article after one or more abrasive operations. The dressing or treatment station may automatically treat or re-dress an abrasive article. For example, based on a next step, a bristle brush or abrasive wheel is treated. A bristle brush may have a surface profile imparted to it. An abrasive wheel may be measured to ensure that a placement position and applied force are accurate.

The dressing or treatment station may include a sensor configured to capture a current state of an abrasive article - for example detecting capping, loading, change in dimensions due to wear, etc. The sensor may be an optical sensor such as a camera, laser, etc. The sensor may also be a weight sensor or force profile sensor. A dressing or treatment protocol may be executed based on sensed information about an abrasive article. Alternatively, in some embodiments, a dressing or treatment protocol is executed based on an operator indication. The treatment station may include a cutter or a cut plate, in some embodiments. Tn some embodiments, the treatment station may include a polycrystalline diamond component, a polycrystalline cubic boron nitride component, or another dressing device. Other treatment and dressing options may also be available.

The dressing / treatment station may be located within a range of motion of a robotic abrading device in some embodiments herein. The dressing / treatment station may have its own controller or receive control signals from a robotic controller.

In some embodiments, a gap detector detects a gap present between an abrasive article and an intended worksurface when a robotic abrading system is in an abrading position. The gap detector may sense a difference between where an expected abrasive surface should be (e.g. in contact with a worksurface) and where it is. The gap detector may include an optical sensor, a laser or another suitable sensing system capable of detecting and measuring a gap. In some embodiments, a robotic arm responsible for moving the abrasive article is positionally-adjusted until the actual abrasive position matches the expected position. Based on the detected difference, information about an actual wear rate since a last gap detection was conducted.

Additionally, based on actual robot coordinates, the gap signal and the relative movements from the shaping I conditioning process, a controller can define the new position as well as recalculate a dimensionally adapted process parameter.

Wear detection and calculations, in some embodiments herein, are done by a system controller, which receives sensor signals from a gap detector and / or dressing station and provides a wear indication that can be used to update process parameters for a robotic operation.

Systems and methods herein allow for accurate conditioning and shaping of abrasive articles using suitable dressing materials such as cutting blades or rotational tools. Abrasive articles used with such systems can be more frequently dressed and, therefore, maintain condition, dimension and shape. Shaping and / or conditioning can be done using a fluid- cooled system to prevent heat-related issues on an abrasive wheel or brush. Systems and methods herein measure actual wear and actual dimensions of abrasive articles, which improves abrading accuracy throughout an operation.

In some embodiments herein, systems include a closed loop process of defining shape, position and condition of abrasive wheels and brushes in robotic applications. FTG. 11 illustrates a method of abrading a component using an automated abrading system in accordance with embodiments herein. Method 1100 describes operation of a self- contained robotic cell. For example, cell 700 described with respect to 700 may operate according to method 1100 in some embodiments. However, other robotic cells, with different components and / or configurations may operate using method 1100 as well.

Method 1100 may be particularly useful for high-precision abrading operations. As described with respect to FIGS. 4A-4C, turbines are time-consuming and labor intensive. Automation, using method 1100, may be useful for automating other precision abrasive operations. For example, cast and forged products such as implants, surgical tools and instruments, golf clubs, plumber hardware, tools, gears and cutlery.

In block 1110, a component to be abraded is imaged. In some embodiments, imaging is done by a robot arm picking up a component and bringing the component to an optical sensor, camera array, topography mapper or other suitable surface topography capturing system. The robot arm may move the component into one or more positions relative to the imaging system, which may move or remain stationary, so that a full 3D surface profile of the component can be constructed. However, in other embodiments, the component to be abraded remains stationary, and the robot arm moves a topography captunng system relative to the component. In yet other embodiments, both the component and the topography capturing system move relative to each other.

In block 1190, an abrasive strategy is generated. A controller, based on the topography captured in block 1110, generates a multi-step abrading strategy. The multi- step abrading strategy may include coarse material removal steps such as cutting or grinding, as well as finer removal steps such as sanding or polishing. As illustrated in method 1 100, the controller generates an abrasive strategy based on machine learning applied to data obtained from previous abrading operations.

In block 1120, a first step of the abrasive strategy is earned out, using a first abrasive article, and material is removed from the component. The target material removal may be, for example, to grind 3 mm of material off the surface.

In block 1130, the component is again imaged. The same imaging system used in block 1110 may be used in block 1130, or a separate imaging system may be used.

In block 1140, the imaging of block 1130 is compared to the imaging in block 1110 and an evaluation is made as to the efficacy of the abrading step of block 1120. For example, while 3 mm was intended to be removed, only 2 8mm was removed. The difference provides a way to estimate a wear rate of the abrasive article. The applied force, movement speed, and dwell time of contact between the abrasive article and the component are known, as well as the difference between the target material removal rate and the actual amount removed. This information allows for calculation of an actual wear level of the abrasive article. All of this information is stored (as noted in block 1150) and machine learning is applied (as illustrated in block 1170) to provide an improved set of parameters for the next abrasive operation in block 1190.

In block 1160, based on the knowledge that only 2.8 mm, for example, instead of 3.0 mm of material was removed, parameter settings of the robotic abrading system are changed. For example, if the next abrading step was to proceed to a finer abrasive to remove 0.4 mm additional material; new parameters are selected so that 0.6 mm of material is removed. Changing parameters may include changing force, dwell time, movement speed, attack angle, abrasive article, lubricant or grinding aids, or any other suitable parameter.

Even when a target material removal rate is achieved (e.g. exactly 3.0 mm of material is removed), settings may be changed in block 1160 to reduce a wear rate of the abrasive article. Imaging, in block 1130, may also capture information about a surface finish. Settings may be changed, in block 1160, to achieve an improved surface finish - such as reduced haze, improved clarity, etc.

As described herein, robotics cell can use imaging of an abrasive surface, the component, or combinations of both, as well as process information from sensors and known process parameters and can correlate these against chronological usage data of the abrasive to understand a state of wear of the abrasive medium. Robotic systems can then take optical, process and environmental information to characterize the best way to utilize an abrasive for maximum life.

System can extend abrasive life via adaptive processing with abrasive hardware to ensure final part is produced. The building of data and deployment of a machine learning model may be done by following the steps of method 1100 over time. As the robotic abrading system performs abrasive tasks, controlled process parameters (such as contact force, feed speed, etc.) and uncontrolled environmental conditions (such as ambient temperature, localized temperature, vibration acceleration and frequency etc.) are recorded. For example, a central cell controller sends commands to a robotic arm and process tools to execute certain toolpaths and abrasive operations, with pre-determined process parameters while active sensing of process parameters and environmental conditions is undertaken, and results are recorded automatically.

For example, a central cell controller sends commands to a robotic arm and to inspection equipment to execute inspection toolpaths and pre-defined inspection and analysis recipes (which may, in turn, executed on the host PCs for the inspection hardware).

Analysis results are returned to the central cell controller, which will (along with process parameters) store these in a central database. This process is repeated until useful abrasive consumable life has been depleted, and then repeat using different parameter sets. A database is built with successive trials, each time becoming broader and deeper in its coverage of process parameter combinations and their effects on process outputs.

FIGS. 12A-12B illustrates a method of evaluating a wear level of an abrasive article in accordance with embodiments herein. FIG. 12A illustrates a method of providing a wear indication.

Observations made on abrasive process outputs can be contextualized with monitoring of process inputs to establish some understanding of the underlying behaviors within a process. For example, the amount of material abraded from a component may be a function of several momtored/controlled process inputs P, environmental factors E, and non-monitored/uncontr oiled process inputs U

Material Removal = f P, E, U) Equation 1

It may be the case that some of these inputs to the function f act independently of others, so can be characterized separately as another function or functions, whereas some may have interactions with other parameters and must be treated as a combined set of inputs to a single function. For example, material removal rate may simply be a function of the contact force F, processing time T, and the belt speed Vf In this case, material removal rate may simply be defined as follows:

Equation 2

Where the terms within the function may describe simple or complex mathematical operations by which the input parameters influence the material removal rate, for example.

(2F + V ) Equation 3

In the case where the material removal rate is influenced by additional factors, the above equation is simply expanded to account for these additional influences. Where additional process parameters (such as remaining abrasive life A, cumulative process time ST, and belt temperature Thai) may influence each other, or may otherwise influence the output of the function these may be expressed as other functions and may also be used in conjunction with primary functions to determine the process outputs:

Material Removal rate = f F, T, V^, g^ Equation 4

Where.

Abrasive life consumption = g(A, T_beit, 2T) Equation 5

The exact form and nature of these relationships may be determined by machine learning techniques, and may not be “human readable. ” E.g. in some embodiments, the deployment of machine learning can result in decision making such that operators see inputs and outputs, but the actual algorithmic decision making process is not transparent. However, it becomes possible to predict the amount of abrasive life that will be consumed either from doing a new process, or that has already been consumed through all historical usage of a particular instance of an abrasive consumable.

For both methodologies, the process of generating “seed” data to build the model, and then deploying the model to predict abrasive life, follows a similar process to that illustrated in FIGS. 12A-12B.

In block 1210, a component is scanned by one or more topography or optical scanners so that a surface contour is obtained.

In block 1220, the amount of material removed is determined by comparing the obtained surface contour to a previous surface contour, obtained before an abrading operation.

In block 1230, the amount of material removed is compared to an expected amount of material removed. For example, if 100 pm of material was expected to be removed, but only 70 pm was, then in block 1240 the force, relative movement and dwell time, environmental and other factors, as well as the actual result are stored.

These parameters are characterized as an operational fingerprint. The only remaining variable is the wear of the abrasive belt, which can then be calculated. In block 1250, an indication of calculated abrasive wear is provided.

These operational fingerprints, stored over the life of an abrasive article, can be used to generate a wear curve for that article, which will help provide detailed understanding of the wear rate for that article and help provide the seed data for a machine learning model to improve abrading efficacy (e g. changing parameters so that, next time, over 95 pm is removed on a first pass) and extend the service life of an abrasive article.

While block 1210 describes scanning the component, it is expressly contemplated that, in some embodiments, the abrasive article can also provide information directly. For example, images of an abrasive belt may be taken, as well as a surface temperature of the belt can also provide insight into the abrasive wear, or understanding as to why an actual abrasive wear deviates from an expected wear - e.g. due to loading or capping. A robotic abrading system may include sensors, optical, thermal or otherwise, that detect surface conditions of an abrasive article.

This approach can also be used to predict abrasive life, by populating the driving knowledge base behind the model with information describing how much of a particular abrasive consumable recorded useful life was consumed by operating an abrasive process with specific process and environmental parameters.

As illustrated in FIG. 12B, based on the collected information from a number of abrasive operations, a tree 1260 may be constructed to determine how different process variables contribute to the process outputs 1270 of material removal rate and abrasive wear rate.

A number of process parameters 1262-1268 are illustrated that may be varied during different experiments to provide the seed data for a machine learning model.

Once enough data has been generated, the decision tree 1260 can be interrogated to provide a prediction of how much abrasive life will be consumed when performing an operation with some known process and environmental parameters. It may be possible with such a dataset for an operator to select to maximize material removal rate, minimize abrasive wear rate, or select a balance between them.

It is expressly contemplated, however, that once a machine learning algorithm has received enough seed data, it continues to further learn as each new operation is conducted and new sensor data is received.

FIG. 13 illustrates a robotic abrading system in accordance with embodiments herein. System 1300 may be an automated robotic abrading cell that conducts an abrasive operation on a workpiece 1312, using an abrasive article 1320. Cell 1300 may operate on a closed circuit feedback loop such that a machine learning model 1346 uses data, stored in datastore 1340, to progressively improve a settings generator 1362. However, while cell 1300 is illustrated as self-contained in FIG. 1300, it is expressly contemplated that, in some embodiments, components may be remote from each other, and communicably coupled using a wired, wireless, or cloud-based network.

A robotic abrading system 1310 includes a robotic arm 1302 that causes interaction between abrasive article 1320 and workpiece 1312. It is expressly contemplated that, in some embodiments, robotic arm 1302 is coupled to abrasive article 1320. However, in other embodiments, robotic arm is coupled to workpiece 1312. In some embodiments, robotic abrading system 1310 includes multiple robotic arms 1302, e.g. one coupled to abrasive article 1320 while a second is coupled to workpiece 1312. Other suitable configurations are expressly contemplated. Robotic abrading system 1310 includes one or more movement mechanisms 1304. Movement mechanisms 1304 may include mechanical j oints that allow robotic arm 1302 to move from a first position to a second position, as well as movement mechanisms 1304 that allow robotic abrading system 1310 to move physically within the cell.

Abrasive article 1320 is a consumable abrasive article that includes particles 1322 that wear or fracture during use, a backing or resin structure is bonded to the abrasive article. Abrasive article 1320 may include other features 1326, such as grinding aids, a second set of particles 1322, etc. Particles 1322, as described herein, may be crushed abrasive particles, formed abrasive particles, shaped abrasive particles or microreplicated abrasive surface features.

Workpiece 1312 has a surface contour 1314 that changes as it is abraded. Workpiece 1312 is composed of a material 1316, or more than one material 1316. Workpiece 1312 may have other features 1318.

A sensor system 1330 may be used to obtain status information of abrasive article 1320 and / or workpiece 1312 during an abrasive process. For example, robotic system 1310 may bring sensor system 1330 to abrasive article 1320 and / or workpiece 1312 in some embodiments. In other embodiments, robotic system 1310 brings abrasive article 1320 and I or workpiece 1312 into position with respect to sensor system 1330. In yet other embodiments, sensor system 1330 is fixed in a position to capture information about abrasive article 1320 and / or workpiece 1312. While only a single sensor system 1330 is illustrated in cell 1300, it is expressively contemplated that multiple sensor systems 1330 may be present within a cell 1300, for example each with a different use Additionally, a single sensor system 1330 may include one or more sensors 1332. Sensor 1332 may be an optical sensor, thermal sensor, or other suitable sensor. Sensor 1332 may include a profilometer, a 2D camera or video camera, a 3D camera or video camera, a camera or video camera array, or another suitable sensor. Sensor 1332 may include a gap sensor that detects a gap between an expected position of an abrasive article and an actual position of the abrasive article.

Sensor system 1330 may include a movement controller 1334 that controls movement of a movement mechanism (not shown) of sensor system 1330, or movement mechanism 1304. A sensor signal communicator 1336 may communicate captured sensor signals to system controller 1350, or to datastore 1340, or to another receiving device. Sensor system 1330 may also include other features 1338. For example, sensor system 1330 may include a light array to improve the quality of captured photos.

Robotic cell 1300 also includes, or is communicably coupled to, datastore 1340. Datastore 1340 includes a number of historical operational fingerprints 1342. Each operational fingerprint 1342 may include some or all of: operational parameters (e.g. force, dwell time, movement speed), environmental conditions (e.g. temperature, humidity), and other abrasive parameters (e.g. abrasive article, lubricant used, grinding aids used), target operational results (e.g. goal abrasive wear rate, goal material removed) as well as operational results (e.g. actual abrasive wear, actual material removal rate). Datastore 1340 may also include job specifications 1344, such as a final surface contour 1314 for workpiece 1312. Job specifications 1344 may also include target abrasive wear rate and / or target operation time (e.g. wear rate of all abrasive steps).

A machine learning model 1346 is illustrated in FIG. 13 as stored in datastore 1340. However, in some embodiments, machine learning model is housed on a processing unit of system controller 1350 and accesses data within datastore 1340. As described herein, machine learning model 1346 may leam from seed data about a number of abrasive operations conducted, and actual wear and material removal rates resulting from known operational parameters. Machine learning model 1346 may then be used to inform settings generator 1362, which selects settings for a next abrasive operation.

Datastore 1340 may also include other information 1348, such as pre and post operation images, surface profiles, renderings, etc. Other information 1348 may also include information about other options that may be used by robotic abrading system, such as other abrasive articles or materials that may be used.

System controller 1350 includes a specifications retriever 1352 which retrieves job specifications 1344, a sensor signal retriever 1354 that retrieves sensor signals directly from system 1330 or retrieves operational fingerprints from datastore 1340. A material removed calculator 1356 may compare a current operational fingerprint 1342 to a previously captured operational fingerprint 1342. A previous surface contour 1314 is compared to a current surface contour 1314 to determine an actual amount of material removed in a last abrasive step.

Based on the actual rate of material removal, and the information from the current operational fingerprint, wear rate calculator 1358 generates an abrasive wear rate, and / or actual abrasive wear of abrasive article 1320.

Settings generator 1362, informed by machine learning model 1346, selects operational parameters and, therefore, operational settings for robotic abrading system 1310, for a next abrasive pass 1362. Settings generator 1362 may determine that a current surface contour 1314 is not sufficient to move to a next abrasive step (e.g. from sanding to polishing, or from grinding to sanding) and may instead indicate that the last abrasive step be repeated with new parameters (e.g. dwell time and force selected based on the calculated wear of the abrasive article, etc.). Settings generator 1362, however, may also determine that the current surface contour 1314 is not a target surface contour, but is sufficient enough to proceed to the next abrasive step, but with adjusted parameters to “catch up” to the desired target surface contour.

Command generator 1364 generates movement instructions and settings adjustment instructions for controllers and components within cell 1300. For example, command generator 1364 may send commands to movement mechanism 1304 and to movement controller 1345 directly, in some embodiments. However, it is expressly contemplated that, in some embodiments, robotic abrading system 1310 has a local controller that receives a command from a command communicator 1366 and causes the command to be implemented locally. Similarly, sensor system(s) 1330 may also have local controllers.

In some embodiments, cell 1300 has a display component 1370 for ahuman operator to view during an operation. Display component 1370 may present a user interface generated by graphical user interface generator 1368 based on information in datastore 1340 or generated by system controller 1350. Display component may present current operational parameters, or settings 1372 of robotic abrading system 1310. Display component may also present a surface profile 1374 of workpiece 1312. This may be, in some embodiments, a surface contour 1314 of a most recent operational fingerprint 1342, or, in other embodiments, a video feed of workpiece 1312 or, in yet other embodiments, a target surface contour set by job specifications 1344. Display component 1370 may also, in some embodiments, display status information 1376 about abrasive article 1320, such as a type, a brand, an expected amount of wear for a current step (e.g. next grinding step will consume 2% of total abrasive life) or an expected remaining service life (e.g. 88% remaining abrasive life, or 56 cycles to next abrasive article replacement). Other information of interest 1378 may also be presented.

Robotic cell 1300 enables the gathering of 3 dimensional geometric and topographical measurement data of a worksurface, through targeted inspection and in- process telemetry, allowing the system controller 1350 to predict the amount of useful work an abrasive consumable can perform. This prediction can be measured in numbers of percentage wear, or total time performing abrasion.

The effect of an abrasive process on a component being abraded can be measured by macro and micro-scale three dimensional (3D) geometric and surface topographical inspection data. Technologies which may be used include: Coordinate measurement machines (CMM); structured light scanning; laser scanning; laser photogrammetry; white light interferometry; confocal microscopy; and focus variation. When used appropriately, each of these technologies can gather accurate point cloud data that provides insights into the target surfaces of a component and detect the effect of both the abrasive process on the component, and the performance of the consumable abrasive material.

The environmental, and abrasive process’, parameters can be measured by in- process telemetry. Sensors may include those suitable for recording (either as analogue voltage or current signals, or encoded digital signals); environmental temperature, humidity, sound pressure levels, localized vibration frequency, accelerations, temperatures, contact forces and tool speeds.

Systems and methods herein provide a more direct assessment of consumable wear, using a machine vision system, which takes digital images of the abrasive article and / or a component being abraded itself during the abrasive process. The vision system can identify changes between subsequent abrading steps and identify stages of wear as a direct result of the recorded process parameters. Through analytics, that information can be correlated to the part quality metrics obtained through geometric and surface topographical inspection. This provides an understanding of the limits of consumable material wear, given the component’s tolerances, and thus the expected lifespan of the consumable for its intended task.

For example, a particular abrasive belt may be selected for performing several consecutive abrasive processes on a component.

E.g. Starting with a fresh belt with no prior usage, several abrasive processes can be performed, during which primary parameters are controlled, and secondary parameters are monitored. The quality of the process outputs are recorded (through geometric and/or topographical inspection) after each individual abrasive process or process step. This provides a summation of all prior usage of the abrasive consumable (for example, cumulative amount of time the belt has been in contact with the component, with weighting of this time as a function of process parameters). Over time, this can be used to understand the conditions required to bring an abrasive consumable to the point where it no longer produces useful results.

This relationship between historical usage and measured quality of outputs can then be used to make minimize or reduce the consumption of “life” on new instances of abrasive consumable (i.e. a replacement belt after the initial one is fully worn) when used in controlled abrasive processes. Using abrasive articles to their full service life, as well as extending the service life of an abrasive article can improve overall efficiency of robotic cell 1300, even if individual steps take longer, as robotic cell 1300 experiences downtime each time an abrasive article needs to be replaced, treated for loading or capping, etc.

Using systems and methods described herein, process parameters can be selected to find a compromise between process quality outputs (such as the amount of material removal and process speed), and consumption of abrasive life.

Systems and methods herein enable visualization of this the cut performance and commensurately life. For abrasive articles without visible grains, it is possible to detect the micro-replicated abrasive block height, which is linked directly life of the product. Process outputs such as acoustics, vibration, and temperature can potentially inform the state of decay of the abrasive surface However, for all abrasive articles, knowing an ‘End of life’ of the article prevents a final product being damaged by an abrasive that is about to run out of life in mid robotic cycle.

FIG. 14 illustrates a networked architecture for a setting selection system for an automated robotic abrading system. FIG. 14 is a networked architecture for a setting selection system 1410. Architecture 1400 illustrates one embodiment of an implementation of a system 1410, however others are possible. In various embodiments, remote servers can deliver the services over a wide area network, such as the internet, using appropriate protocols. For instance, remote servers can deliver applications over a wide area network and they can be accessed through a web browser or any other computing component.

Software or components, as well as the corresponding data, can be stored on servers at a remote location. The computing resources in a remote server environment can be consolidated at a remote data center location or they can be dispersed. Remote server infrastructures can deliver services through shared data centers, even though they appear as a single point of access for the user. Thus, the components and functions described herein can be provided from a remote server at a remote location using a remote server architecture. Alternatively, they can be provided by a conventional server, installed on client devices directly, or in other ways.

As described herein, setting selection system 1410 selects operational settings for components of a robotic abrading system based on information received from one or more sensors 1480, which may detect information about an abrasive article, workpiece, or other operational information about robotic abrading system 1470. As illustrated, setting selection system 1410 may communicate directly with settings 1480 and robotic abrading system 1470, in some embodiments herein.

FIG. 14 specifically shows that a system 1410 can be located at a remote server location 1402. Therefore, computing device 1420 accesses those systems through remote server location 1402. Operator 1450 can use computing device 1420 to access user interfaces 1422 as well. For example, user interface 1422 may provide an indication of how worn an abrasive article is, changes that are made to any of networked systems 1404, or suggested changes to the operation by the operator - such as increasing force, increasing RPMs, etc.

FIG. 14 shows that it is also contemplated that some elements of systems described herein are disposed at remote server location 1402 while others are not. By way of example, storage 1430, 1440 or 1460 or robotic abrading system 1470 can be disposed at a location separate from location 1402 and accessed through the remote server at location 1402. Regardless of where they are located, they can be accessed directly by computing device 1420, through a network (either a wide area network or a local area network), hosted at a remote site by a service, provided as a service, or accessed by a connection service that resides in a remote location. Also, the data can be stored in substantially any location and intermittently accessed by, or forwarded to, interested parties. For instance, physical carriers can be used instead of, or in addition to, electromagnetic wave carriers.

It will also be noted that the elements of systems described herein, or portions of them, can be disposed on a wide variety of different devices. Some of those devices include servers, desktop computers, laptop computers, imbedded computer, industrial controllers, tablet computers, or other mobile devices, such as palm top computers, cell phones, smart phones, multimedia players, personal digital assistants, etc.

FIGS. 15 illustrates an example computing device that can be used in embodiments shown in previous Figures. FIG. 15 is one example of a computing environment in which elements of systems and methods described herein, or parts of them (for example), can be deployed. With reference to FIG. 15, an example system for implementing some embodiments includes a general -purpose computing device in the form of a computer 1510. Components of computer 1510 may include, but are not limited to, a processing unit 1520 (which can comprise a processor), a system memory 1530, and a system bus 1521 that couples various system components including the system memory to the processing unit 1520. The system bus 1521 may be any of several types of bus structures including a memory bus or memory controller, a peripheral bus, and a local bus using any of a variety of bus architectures Memory and programs described with respect to systems and methods described herein can be deployed in corresponding portions of FIG. 15.

Computer 1510 typically includes a variety of computer readable media. Computer readable media can be any available media that can be accessed by computer 1510 and includes both volatile/nonvolatile media and removable/non-removable media. By way of example, and not limitation, computer readable media may comprise computer storage media and communication media. Computer storage media is different from, and does not include, a modulated data signal or carrier wave. It includes hardware storage media including both volatile/nonvolatile and removable/non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by computer 1510. Communication media may embody computer readable instructions, data structures, program modules or other data in a transport mechanism and includes any information delivery media. The term “modulated data signal” means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal.

The system memory 1530 includes computer storage media in the form of volatile and/or nonvolatile memory such as read only memory (ROM) 1531 and random access memory (RAM) 1532. A basic input/output system 1533 (BIOS) containing the basic routines that help to transfer information between elements within computer 1510, such as during start-up, is ty pically stored in ROM 1531. RAM 1532 typically contains data and/or program modules that are immediately accessible to and/or presently being operated on by processing unit 1520. By way of example, and not limitation, FIG. 15 illustrates operating system 1534, application programs 1535, other program modules 1536, and program data 1537.

The computer 1510 may also include other removable/non-removable and volatile/nonvolatile computer storage media. By way of example only, FIG. 15 illustrates a hard disk drive 1541 that reads from or writes to non-removable, nonvolatile magnetic media, nonvolatile magnetic disk 1552, an optical disk drive 1555, and nonvolatile optical disk 1556 The hard disk drive 1541 is typically connected to the system bus 1521 through a non-removable memory interface such as interface 1540, and optical disk drive 1555 are typically connected to the system bus 1521 by a removable memory interface, such as interface 1550.

Alternatively, or in addition, the functionality described herein can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that can be used include Field- programmable Gate Arrays (FPGAs), Application-specific Integrated Circuits (e.g., ASICs), Application-specific Standard Products (e.g., ASSPs), System-on-a-chip systems (SOCs), Complex Programmable Logic Devices (CPLDs), etc The drives and their associated computer storage media discussed above and illustrated in FIG. 15, provide storage of computer readable instructions, data structures, program modules and other data for the computer 1510. In FIG. 15, for example, hard disk drive 1541 is illustrated as storing operating system 1544, application programs 1545, other program modules 1546, and program data 1547. Note that these components can either be the same as or different from operating system 1534, application programs 1535, other program modules 1536, and program data 1537.

A user may enter commands and information into the computer 1510 through input devices such as a keyboard 1562, a microphone 1563, and a pointing device 1561, such as a mouse, trackball or touch pad. Other input devices (not shown) may include a joystick, game pad, satellite receiver, scanner, or the like. These and other input devices are often connected to the processing unit 1520 through a user input interface 1560 that is coupled to the system bus, but may be connected by other interface and bus structures. A visual display 1591 or other type of display device is also connected to the system bus 1521 via an interface, such as a video interface 1590. In addition to the monitor, computers may also include other peripheral output devices such as speakers 1597 and printer 1596, which may be connected through an output peripheral interface 1595.

The computer 1510 is operated in a networked environment using logical connections, such as a Local Area Network (LAN) or Wide Area Network (WAN) to one or more remote computers, such as a remote computer 1580.

When used in a LAN networking environment, the computer 1510 is connected to the LAN 1571 through a network interface or adapter 1570. When used in a WAN networking environment, the computer 1510 typically includes a modem 1572 or other means for establishing communications over the WAN 1573, such as the Internet. In a networked environment, program modules may be stored in a remote memory storage device. FIG. 15 illustrates, for example, that remote application programs 1585 can reside on remote computer 1580.

By gathering macro and micro-scale 3 dimensional geometric and topographical measurement data, through targeted inspection and in-process telemetry, systems and method herein can be used to extend the amount of useful work an abrasive consumable can perform. This extension in work is typically measured in numbers of components processed or total time performing abrasion Geometric and topographical inspection technologies which may be used include structured light scanning, laser photogrammetry, white light interferometry, confocal microscopy, and focus variation) and in-process telemetry sensors may include those suitable for recording (either as analogue voltage or current signals, or encoded digital signals): environmental temperature, humidity, sound pressure levels, and localized vibration frequency and accelerations, temperatures, and tool speeds.

By performing several controlled abrasive processes, and monitoring both primary process parameters (belt surface speed, belt feed speed, contact force, abrasive grain type and size), and secondary process parameters (belt/component surface temperatures, ambient temperatures, vibration, and noise levels etc ), the performance of the abrasive consumable (measured by its effectiveness in producing desired results on the component being processed) can be correlated to process parameters.

Then, by forming an understanding of how well an abrasive consumable performs in achieving desired process results (geometric or topographical) on a component, in the context of all the historical usage which that abrasive consumable article has seen, an approximation of the percentage of useful abrasive “life” consumed as a function of a range of process parameters can be determined.

Objects and advantages of this disclosure are further illustrated by the following non-limiting examples, but the particular materials and amounts thereof recited in these examples, as well as other conditions and details, should not be construed to unduly limit this disclosure.

An automated abrasive article evaluation system is presented that includes a sensor signal retriever that retrieves a pre-operation signal and a post-operation signal for an abrasive operation. The system also includes a specification retriever that retrieves a set of operational parameters for the abrasive operation. The system also includes a material removed calculator that compares the pre-operational signal to the post-operational signal and determines a material removed amount. The system also includes a wear rate calculator that, based on the material removed amount and the set of operational parameters, calculates an amount of wear of an abrasive article used in the abrasive operation. The system also includes a wear communicator that communicates the amount of wear to a non-transitory storage medium. The system may be implemented such that the pre-operation signal is a pre-operation image, and wherein the post-operation signal is a post-operation image.

The system may be implemented such that the pre-operation signal comprises a preoperation topography and wherein the post-operation signal is a post-operation topography.

The system may be implemented such that the pre-operation signal comprises a first workpiece surface contour before the abrasive operation, and wherein the post-operation signal comprises a second workpiece surface contour after the abrasive operation.

The system may be implemented such that the pre-operation signal comprises a first abrasive surface topography, of the abrasive article, before the abrasive operation, and wherein the post-operation signal comprises a second abrasive article topography.

The system may be implemented such that it includes a job specification retriever that retrieves a target workpiece surface contour and compares the second workpiece surface contour to the target workpiece surface contour and, based on the comparison, provides a deviation indication.

The system may be implemented such that the wear communicator communicates the amount of wear to a display component.

The system may be implemented such that the wear communicator communicates the amount of wear to a controller of a robotic abrading system associated with the abrading operation.

A robotic abrading system is presented that includes a robotic arm that causes a workpiece to contact an abrasive article according to a set of operational parameters for an abrasive operation. The system also includes a pre-operation topography capturing device that captures a pre-operation surface topography for a portion of the workpiece. The system also includes a post-operation topography capturing device that captures a post-operation surface topography for the portion of the workpiece. The system also includes a wear calculator that calculates a wear level of the abrasive article after the abrasive operation. The system also includes a controller that generates an abrasive strategy based on the preoperation surface topography, the abrasive strategy comprising the set of operational parameters, and the calculated wear level.

The robotic abrading system may be implemented such that the pre-operation topography captunng device is the post-operation topography capturing device. The robotic abrading system may be implemented such that the robot removably couples to the workpiece.

The robotic abrading system may be implemented such that the robot arm moves along a movement path to bring the workpiece from a pickup location to a capture location. The capture location comprises the workpiece in a capture position with respect to the preoperation topography capturing device.

The robotic abrading system may be implemented such that the controller generates the abrasive strategy in situ, such that the robotic arm conducts the abrasive operation after the abrasive strategy is generated.

The robotic abrading system may be implemented such that the robotic arm moves along a second movement path from an abrading location, after the abrasive operation, to a post-operation capture location. The post-operation capture location comprises the workpiece in a post-capture position with respect to the post-operation topography capturing device.

The robotic abrading system may be implemented such that it includes a material removal calculator that compares the post-operation surface topography to the pre-operation topography and, based on the comparison, calculates an actual material removed.

The robotic abrading system may be implemented such that it includes an abrasive operation evaluator that retrieves a target surface topography and compares the postoperation surface topography to the target surface topography and generates a deviation indication based on the comparison.

The robotic abrading system may be implemented such that the controller generates the abrasive strategy based on the deviation indication.

The robotic abrading system may be implemented such that the abrading operation is a first abrading operation in an abrading sequence, and, for a second abrading operation in the abrading sequence, the controller retrieves a template abrasive strategy' for the second abrading operation and generates a second abrasive strategy by modifying the template abrasive strategy based on the deviation indication and the wear level.

The robotic abrading system may be implemented such that the pre-operation topography capturing device comprises a camera array, a profilometer, a coordinate measurement machine, a structured light scanner, a laser scanner, a confocal microscope or a white light interferometer. A automated abrading system is presented that includes a robot arm configured to cause an abrasive article to contact a workpiece for an abrading operation. The abrading operation comprises contact between the abrasive article and the workpiece at an applied force, for a dwell time, at a movement speed. The system also includes a sensor system that captures a first sensor signal before the abrading operation and a second sensor signal after the abrading operation. The system also includes an article evaluation system that: receives the first and second sensor signals, receives the applied force, the dwell time and the movement speed, and based on the received sensor signals and the received applied force, dwell time and movement speed, calculates a wear level of the abrasive article. The system also includes a robotic controller that, based on the calculated wear indication and the second sensor signal, generates a second applied force, a second dwell time and a second movement speed for a second abrading operation.

The system may be implemented such that the robotic controller generates the second applied force, second dwell time and second movement speed by: retrieving a target surface topography for the workpiece for the abrading operation, using the second sensor signal, generating an actual surface topography for the workpiece, comparing the target surface topography to the actual surface topography, detecting a deviation between the target surface topography and the actual surface topography, and modifying a planned second applied force, a planned second dwell time and a planned second movement speed to account for the deviation.

The system may be implemented such that the abrading operation is a first abrading step in an abrading sequence, and further comprising: the robot controller retrieving a set of operational parameters for a second abrading step and, based on the calculated wear level, modifying the set of operational parameters to the second applied force, the second dwell time and the second movement speed.

The system may be implemented such that the applied force, the dwell time, the movement speed, the first sensor signal and the second sensor signal are stored in a datastore as a first operational fingerprint.

The system may be implemented such that the datastore comprises a plurality of historic operational fingerprints, and the robotic controller comprises a machine learning model that, based on the plurality of histone operational fingerpnnts, generates the second applied force, the second dwell time and the second movement speed. The system may be implemented such that the first sensor signal comprises an image of the abrasive article.

The system may be implemented such that the first sensor signal comprises an image of the workpiece.

The system may be implemented such that the first sensor signal comprises a topography of the workpiece.

The system may be implemented such that the first sensor signal comprises a rendering of the workpiece.

The system may be implemented such that the abrasive article is a coated abrasive belt.

The system may be implemented such that the robot arm is coupled to the workpiece.

The system may be implemented such that the robot arm is coupled to the abrasive article.

The system may be implemented such that the robot arm comprises a force control unit.

The system may be implemented such that the abrasive article is an abrasive belt that moves at the movement speed.

The system may be implemented such that the abrasive article is a grinding wheel that rotates at the movement speed.

The system may be implemented such that the abrasive article is an abrasive disc that rotates at the movement speed.

The system may be implemented such that the sensor signal is received from an optical sensor

The system may be implemented such that the sensor signal is received from a profilometer.

A method of abrading a substrate is presented that includes capturing, using a topography capturing device, a surface profile of the substrate. The method also includes generating, based in part on the captured surface profile, a wear level of an abrasive article. The method also includes generating an abrasive strategy, based on the surface profile and the wear level, for abrading the substrate from the surface profile to a target profile, using the abrasive article. The method also includes causing a robotic abrading system to execute the abrasive strategy in an abrasive operation. The method may be implemented such that it includes capturing a post-operation surface profde of the substrate, after the robotic abrading system completes the abrasive operation, comparing the post-operation surface profde of the substrate to the target-profde, and generating a deviation indication based on the comparison.

The method may be implemented such that it includes comparing the deviation indication to a threshold and, if the deviation indication exceeds the threshold, causing the robotic abrading system to execute a second abrasive strategy with the abrasive article.

The method may be implemented such that the abrasive operation is a first abrasive operation in a sequence of abrasive operations and further includes comparing the deviation indication to a threshold and, if the deviation indication is less than a threshold, causing the controller to generate a second abrasive strategy, for a second abrasive operation, based on the post-operation surface profile.

The method may be implemented such that it also includes generating a second wear level based on the post-operation surface profile. The second abrasive strategy is based on the second wear level.

The method may be implemented such that the topography capturing device comprises a camera array, a profilometer, a coordinate measurement machine, a structured light scanner, a laser scanner, a confocal microscope or a white light interferometer.

The method may be implemented such that the robotic abrading system comprises a robotic arm that picks up the substrate and causes the substrate to contact the abrasive article at an applied force for a contact time. The abrasive strategy comprises the applied force and the contact time.

The method may be implemented such that the robotic abrading system comprises a robotic arm coupled to an abrasive article.

EXAMPLES

EXAMPLE 1

FIG. 16 illustrates a example sequence of abrasive operations that can be done with a robotic abrading unit to cut a target profile 1650 shape out of a starting workpiece 1610. While only two steps are illustrated, it is noted that this may take more steps in practice.

A first grinding pass 1615 is programmed and executed with a target goal of removing a large quantity of material around the top curve. After the first pass, imaging is done to generate the image 1620 of the first pass result. Using systems and methods herein, the second grinding step 1625 can be modified based on the results of the first grinding step 1615. For example, the first grinding step 1615 could have removed more material - this indicates that the abrasive article was more worn than expected, resulting in a lower cut rate. Second grinding step 1625 can now be adjusted to remove more material so that the result after the the second pass 1630 are in line with what was originally expected after two passes.

After the second pass, the system evaluates second pass result 1630 and may determine that, since it is close to a target profile 1650, that it can move to a polishing sequence, which can be modified to directly target areas that still have more material removal needs.

EXAMPLE 2

Figure 17A shows the decay of abrasive belts when subjected to multiple material removal cycles. This decay is measured in the reduced ability of the abrasive belt to removal material(17A), an increase in grinding temperature (17B), and a change to the surface topography of the substrate being abraded (17C).

The equipment used was a 60 horsepower back stand machine which was running with a spindle speed of 2000Rpm. This spindle drove a 12” polyurethane coated contact wheel with a 95° shore hardness, and a 1 :2 land to groove ratio. This equipment was run with a selection P36 and 36+ abrasives belts which were 3500mm in length and 100mm in width.

With the spindle speed of 2000RPM and use of a 12” 0 contact wheel, the resultant surface speed for the abrasives was 2000 x ((Pi x 0.3))/60 = 31m/s. This surface speed was used as a constant for all experiments.

The abrasion method employed was a plunge test, where a 2Omm0 rod of stainless steel 304L was weighed and then pushed/plunged into an abrasive belt being by driven by the contact wheel, at a constant force against for a constant amount of time. The constant force of 170N (0.45N/mm2) was utilized for all experiments. Additionally, in all cases a constant plunge time of 20 seconds was used and repeated on brand-new stainless-steel bars to ensure that there was no residual heat from previous plunges. 60 plunge tests were conducted at a constant of 150N to assess the abrasive decay and ability to remove material. This was measured by weighing the stainless-steel rod before grinding, and then weighing the rod after grinding to calculate a delta weight which then allowed calculation of the material removal. Surface finish measurements were taken were a profilometer at regular intervals and substrate temperatures were measured directly after griding to assess the impact of the abrasive decay on the thermal characteristics of the substrate.

The results observed were that in subtractive manufacturing applications such as material removal with abrasive belts, the abrasive wear is directly related to the rate of material removal, in that with each additional cycle of abrasion material removal decreases in fixed operating conditions. Commensurately temperature increases due to the dulling / fracturing of the mineral tips creating more friction, and surface finish (scratch depth) was reduced due to a shallower impregnation of the mineral into the substrate. As it is shown in FIG 17D, it appears that there is a correlation between material removal and surface temperature, more and less independent of the type of the abrasive belt product. This clearly indicates that for a given condition, temperature sensing during the abrasive process can be used as a live input data for material removal and/or abrasive belt decay estimation. Thus, it helps to adjust other parameters such as speed or force in order to maintain the material removal or extend the life of the abrasive belt.

Claims

What is claimed is:

1. An automated abrasive article evaluation system comprising: a sensor signal retriever that retrieves a pre-operation signal and a post-operation signal for an abrasive operation; a specification retriever that retrieves a set of operational parameters for the abrasive operation; a material removed calculator that compares the pre-operational signal to the post- operational signal and determines a material removed amount; a wear rate calculator that, based on the material removed amount and the set of operational parameters, calculates an amount of wear of an abrasive article used in the abrasive operation; and a wear communicator that communicates the amount of wear to a non-transitory storage medium.

2. The system of claim 1, wherein the pre-operational signal or the post-operational signal comprises a detected gap between the abrasive article and a work-surface.

3. The system of claim 1 or 2, wherein the pre-operation signal is a pre-operation image, and wherein the post-operation signal is a post-operation image.

4. The system of any of claims 1-3, wherein the pre-operation signal comprises a preoperation topography and wherein the post-operation signal is a post-operation topography.

5. The system of any of claims 1-4, wherein the pre-operation signal comprises a first workpiece surface contour before the abrasive operation, and wherein the post-operation signal comprises a second workpiece surface contour after the abrasive operation.

6. The system of any of claims 1-5, wherein the pre-operation signal comprises a first abrasive surface topography, of the abrasive article, before the abrasive operation, and wherein the post-operation signal comprises a second abrasive article topography.

7. The system of claim 4, and further comprising: a job specification retriever that retrieves a target workpiece surface contour and compares the second workpiece surface contour to the target workpiece surface contour and, based on the comparison, provides a deviation indication.

8. The system of any of claims 1-7, wherein the wear communicator communicates the amount of wear to a display component.

9. The system of any of claims 1-8, wherein the wear communicator communicates the amount of wear to a controller of a robotic abrading system associated with the abrading operation.

10. The system of any of claims 1-9, wherein the abrasive article comprises a coated abrasive article, a bonded abrasive article or a nonwoven abrasive article.

11. A robotic abrading system that comprises: a robotic arm that causes a workpiece to contact an abrasive article according to a set of operational parameters for an abrasive operation; a pre-operation topography capturing device that captures a pre-operation surface topography for a portion of the workpiece; a post-operation topography capturing device that captures a post-operation surface topography for the portion of the workpiece; a wear calculator that calculates a wear level of the abrasive article after the abrasive operation; and a controller that generates an abrasive strategy based on the pre-operation surface topography, the abrasive strategy comprising the set of operational parameters, and the calculated wear level.

12. The robotic abrading system of claim 12, wherein the pre-operation topography capturing device is the post-operation topography capturing device.

13. The robotic abrading system of claim 11 or 12, wherein the robot removably couples to the workpiece.

14. The robotic abrading system of claim 13, wherein the robot arm moves along a movement path to bring the workpiece from a pickup location to a capture location, wherein the capture location comprises the workpiece in a capture position with respect to the pre-operation topography capturing device.

15. The robotic abrading system of claim 14, wherein the controller generates the abrasive strategy in situ, such that the robotic arm conducts the abrasive operation after the abrasive strategy is generated.

16. The robotic abrading system of any of claims 11-15, wherein the robotic arm moves along a second movement path from an abrading location, after the abrasive operation, to a post-operation capture location, wherein the post-operation capture location comprises the workpiece in a post-capture position with respect to the post-operation topography capturing device.

17. The robotic abrading system of any of claims 11-16, and further comprising: a material removal calculator that compares the post-operation surface topography to the pre-operation topography and, based on the comparison, calculates an actual material removed.

18. The robotic abrading system of any of claims 11-17, and further comprising: an abrasive operation evaluator that retrieves a target surface topography and compares the post-operation surface topography to the target surface topography and generates a deviation indication based on the comparison.

19. The robotic abrading system of claim 18, wherein the controller generates the abrasive strategy based on the deviation indication.

20. The robotic abrading system of claim 18, wherein the abrading operation is a first abrading operation in an abrading sequence, and wherein, for a second abrading operation in the abrading sequence, the controller retrieves a template abrasive strategy for the second abrading operation and generates a second abrasive strategy by modifying the template abrasive strategy based on the deviation indication and the wear level.

21. The robotic abrading system of claim 18, wherein the pre-operation topography capturing device comprises a camera array, a profilometer, a coordinate measurement machine, a structured light scanner, a laser scanner, a confocal microscope or a white light interferometer.

22. A automated abrading system comprising: a robot arm configured to cause an abrasive article to contact a workpiece for an abrading operation, wherein the abrading operation comprises contact between the abrasive article and the workpiece at an applied force, for a dwell time, at a movement speed: a sensor system that captures a first sensor signal before the abrading operation and a second sensor signal after the abrading operation; an article evaluation system that: receives the first and second sensor signals; receives the applied force, the dwell time and the movement speed; and based on the received sensor signals and the received applied force, dwell time and movement speed, calculates a wear level of the abrasive article; and a robotic controller that, based on the calculated wear indication and the second sensor signal, generates a second applied force, a second dwell time and a second movement speed for a second abrading operation.

23. The system of claim 22, wherein the robotic controller generates the second applied force, second dwell time and second movement speed by: retrieving a target surface topography for the workpiece for the abrading operation; using the second sensor signal, generating an actual surface topography for the workpiece; comparing the target surface topography to the actual surface topography; detecting a deviation between the target surface topography and the actual surface topography; and modifies a planned second applied force, a planned second dwell time and a planned second movement speed to account for the deviation.

24. The system of claim 23, wherein the abrading operation is a first abrading step in an abrading sequence, and further comprising: the robot controller retrieving a set of operational parameters for a second abrading step and, based on the calculated wear level, modifies the set of operational parameters to the second applied force, the second dwell time and the second movement speed.

25. The system of claim 22, wherein the applied force, the dwell time, the movement speed, the first sensor signal and the second sensor signal are stored in a datastore as a first operational fingerprint.

26. The system of claim 25, wherein the datastore comprises a plurality of historic operational fingerprints, and wherein the robotic controller comprises a machine learning model that, based on the plurality of historic operational fingerprints, generates the second applied force, the second dwell time and the second movement speed.

27. The system of any of claims 22-26, wherein the first sensor signal comprises an image of the abrasive article.

28. The system of any of claims 22-26, wherein the first sensor signal comprises an image of the workpiece.

29. The system of any of claims 22-26, wherein the first sensor signal comprises a topography of the workpiece.

30. The system of any of claims 22-26, wherein the first sensor signal comprises a rendering of the workpiece.

31. The system of any of claims 22-30, wherein the abrasive article is a coated abrasive belt.

32. The system of any of claims 22-31, wherein the robot arm is coupled to the workpiece.

33. The system of any of claims 22-32, wherein the robot arm is coupled to the abrasive article.

34. The system of any of claims 22-33, wherein the robot arm comprises a force control unit.

35. The system of any of claims 22-34, wherein the abrasive article is an abrasive belt that moves at the movement speed.

36. The system of any of claims 22-30, wherein the abrasive article is a grinding wheel that rotates at the movement speed.

37. The system of any of claims 22-30, wherein the abrasive article is an abrasive disc that rotates at the movement speed.

38. The system of any of claims 22-37, wherein the sensor signal is received from an optical sensor.

39. The system of any of claims 22-37, wherein the sensor signal is received from a profilometer.

40. A method of abrading a substrate, the method comprising: capturing, using a topography capturing device, a surface profile of the substrate; generating, based in part on the captured surface profile, a wear level of an abrasive article; generating an abrasive strategy, based on the surface profile and the wear level, for abrading the substrate from the surface profile to a target profile, using the abrasive article; and causing a robotic abrading system to execute the abrasive strategy in an abrasive operation.

41. The method of claim 40, and further comprising: capturing a post-operation surface profile of the substrate, after the robotic abrading system completes the abrasive operation; comparing the post-operation surface profile of the substrate to the target-profile; and generating a deviation indication based on the comparison.

42. The method of claim 41, and further comprising: comparing the deviation indication to a threshold and, if the deviation indication exceeds the threshold, causing the robotic abrading system to execute a second abrasive strategy with the abrasive article.

43. The method of claim 41, wherein the abrasive operation is a first abrasive operation in a sequence of abrasive operations and further comprising: comparing the deviation indication to a threshold and, if the deviation indication is less than a threshold, causing the controller to generate a second abrasive strategy, for a second abrasive operation, based on the post-operation surface profile.

44. The method of claim 43, and further comprising, generating a second wear level based on the post-operation surface profile, and wherein the second abrasive strategy is based on the second wear level.

45. The method of any of claims 40-44, wherein the topography capturing device comprises a camera array, a profilometer, a coordinate measurement machine, a structured light scanner, a laser scanner, a confocal microscope or a white light interferometer.

46. The method of any of claims 40-45, wherein the robotic abrading system comprises a robotic arm that picks up the substrate and causes the substrate to contact the abrasive article at an applied force for a contact time, wherein the abrasive strategy comprises the applied force and the contact time.

47. The method of any of claims 40-46, wherein the robotic abrading system comprises a robotic arm coupled to an abrasive article.

48. A robotic system for abrading a surface, the system comprising: a robotic arm configured to move an abrasive article to an abrading position; a wear indicator configured to generate a sensor signal related to a wear level of the abrasive article; a dressing station configured to address a condition of the abrasive article after a first abrasive operation and before a second abrasive operation; and an abrading controller configured to generate a set of coordinates for the robotic arm, the set of coordinates corresponding to the abrading position.

49. The system of claim 48, wherein the abrasive article comprises a bonded abrasive article, a coated abrasive article or a nonwoven abrasive article.

50. The system of claim 48, and further comprising a gap detector configured to detect a gap between the abrasive article and a worksurface, wherein the sensor signal comprises the detected gap.

51. The system of claim 50, wherein the gap detector comprises a camera.

52. The system of claim 50, wherein the abrading controller generates the set of coordinates based on the detected gap.

53. The system of claim 48, wherein the condition comprises capping and wherein dressing the condition comprises a capping removal process. The system of claim 48, wherein the condition comprises loading and wherein addressing the condition comprises de-loading the abrasive article. The system of claim 48, wherein the condition comprises a surface profile of the abrasive article and wherein addressing the condition comprises changing the surface profile.