DE102020006436A1 - Self-propelled gap measuring unit - Google Patents

Abstract

The invention relates to a self-propelled gap measuring unit (1), comprising a mobile platform (2), a robot (3), a sensor head (4) and a computer unit, the robot (3) being attached to the mobile platform (2) and a Has arm with several joints, the sensor head (4) being arranged at the end of the arm of the robot (3), the sensor head (4) having two triangulation sensors, a camera (5) designed as a mono camera system also being provided is, wherein the camera (5) is attached to a boom which is also attached to the mobile platform (2), wherein the mobile platform (2) has two laser scanners which are configured for cooperation according to a SLAM algorithm, the Triangulation sensors are configured to work with different wavelengths, the camera (5) being configured to carry out referencing by means of edge-based tracking, with edge-based tracking Edges of a body recognized in an image field of the camera are compared with edges of a stored CAD model.

Description

The invention relates to a self-propelled gap measuring unit according to claim 1.

It is known to measure body features such as gap and transition between add-on parts with several permanently installed robots. For the referencing, a camera system is used that is aimed at several characteristic features on the vehicle. The referencing provides a position relative to the measured body for the current vehicle.

The DE 10 2009 039 812 A1 discloses a mobile measuring device comprising a carrier vehicle having wheels and drives for moving the wheels, a multi-axis robot arm of a measuring robot attached to the carrier vehicle and a measuring body attached to the carrier vehicle for calibrating the measuring robot.

The invention is based on the object of specifying an improved mobile measuring device.

The object is achieved according to the invention by a self-propelled gap measuring unit according to claim 1.

Advantageous refinements of the invention are the subject matter of the subclaims.

A self-propelled gap measuring unit according to the invention comprises a mobile platform, a robot, a sensor head and a computer unit, the robot being attached to the mobile platform and having an arm with several joints, the sensor head being arranged at the end of the arm of the robot. According to the invention, the sensor head has two triangulation sensors, and a camera designed as a mono camera system is also provided, which is attached to a boom that is also attached to the mobile platform, the mobile platform having two laser scanners that are used for cooperation are configured according to a SLAM algorithm, the triangulation sensors being configured to work with different wavelengths, the camera being configured to carry out referencing by means of edge-based tracking, with edges of a body with edges detected in an image field of the camera during edge-based tracking a stored CAD model can be compared.

The self-propelled gap measuring unit according to the invention allows gap and transition measurements, comprises a robot system, is modular, fast, highly accurate, autonomous and independent of the environment.

The present invention allows a vehicle to be measured quickly in a flexible manner without long planning and / or construction phases. Compared to known solutions, environmental requirements and space requirements, power consumption, set-up effort, effort of integration into an existing line and costs are reduced.

The self-propelled gap measuring unit according to the invention is independent and has no special requirements for the periphery, since everything is available in the platform. The gap measuring unit can be set up and put into operation quickly, whether on the production line or in test boxes outside of the conveyor system. All components of the gap measuring unit are modular and can be replaced by other modules if this is necessary for the fulfillment of the task. For example, a larger robot can be installed at any time as long as it uses a standardized interface (e.g. ROS, Java, etc.), if this is necessary to carry out the measurement task. Other sensors that can also measure on glass, for example, can also be integrated into the measurement software. The accuracy of the self-propelled gap measuring unit is very high. To measure a complete car, depending on the number of points, it takes between five minutes and nine minutes, for example. This system revolutionizes gap measurement and can be used almost anywhere. Thanks to quick commissioning, the production process can be checked and verified at short notice. A manual measuring station can also be converted seamlessly into a gap measuring unit.

In addition, customer vehicles can also be checked, for example in a kind of parking lot. The bodies could be measured during the night or during shift breaks and further processed or transported without loss of time for the employee.

Embodiments of the invention are explained in more detail below with reference to drawings.

• 1 eine schematische Ansicht einer selbstfahrenden Spaltmesseinheit, und
• 2 schematisch einen Messablauf sowie auftretende Varianzen.

Show:

• 1 a schematic view of a self-propelled gap measuring unit, and
• 2 a schematic of a measurement process and any variances that occur.

Corresponding parts are provided with the same reference symbols in all figures.

1 Figure 3 is a schematic view of a self-propelled gap measuring unit 1 , which is designed as a mobile unit. The self-propelled gap measuring unit 1 includes a mobile platform 2 , a robot 3 , in particular a lightweight robot and a sensor head 4th with two triangulation sensors and a camera 5 , for example a mono camera system. The self-propelled gap measuring unit 1 is therefore very space-saving and, thanks to the ability to cooperate between humans and robots, can do without high protective fences. A built-in computer unit (not shown) enables the self-propelled gap measuring unit 1 determine the gap transition without periphery.

The robot 3 is on the mobile platform 2 and has an arm with multiple joints. The sensor head 4th is at the end of the robot's arm 3 arranged. The camera is attached to a boom that is also attached to the mobile platform 2 is attached.

Due to the modular structure, individual components of the self-propelled gap measuring unit 1 replaced in order to meet the requirements. Due to the modularity, specific processes and their requirements can also be met.

In the 1 illustrated mobile platform 2 a Kuka Motion Platform can be KMP for short. Two sensors, in particular laser scanners, which cooperate according to a SLAM algorithm (Simultaneous Localization and Mapping) can be used for navigation. For this purpose, a map is recorded with the sensors, which the mobile platform uses 2 can then position in space. The mobile platform 2 can for example be programmed with Java in the Kuka Sunrise environment. The positioning accuracy of the platform can be σ _{KMP = 10mm, for example.}

The robot 3 can be designed as a lightweight robot. For example, an LBR iiwa 14R lightweight robot from KUKA can be used. It can be coupled and programmed with the Kuka Motion Platform without any problems, as both work with the Java-based Kuka Sunrise. For robots 3 can have a repeat _{accuracy of σ LBR} = 0.3mm.

The sensor head 4th can be used as a double head sensor at the end of the robot 3 be arranged. For example, the sensor head includes 4th two sensors, in particular an MLSL122 and an MLSL132 from Wenglor. Both sensors work preferably according to the triangulation principle, but with different wavelengths, so that every opaque surface can be measured synchronously and independently of color. The accuracy of the sensors can be, for example, σ _DKS = 0.001 mm.

The camera 5 is used for referencing and can be designed as a mono camera system, for example from Daimler Protics GmbH. The referencing principle is based on edge-based tracking. The edges of the body currently recognized in the image field are compared with those of a stored CAD model and the 6D transformation (3x translation, 3x rotation) from camera to body is determined from this. The platform can be positioned by shifting the CAD model relative to the measured body. The accuracy of the camera can be σ _MKS = 3mm.

The computer unit, not shown, can use measurement software from VMT, for example, which is also used in other measurement systems. Therefore, the measured values can be compared and integrated into existing data storage and evaluation systems. In addition to the measurement software, software from the camera system can also run on the computer unit.

The programming and setup of the measuring points is done intuitively using a software and control package, the Daimler TSS. This is done by the robot 3 guided by hand to the points to be measured. A test procedure checks all routes and points before the automated measurement. Referencing on the basis of the virtual gap points makes variances when approaching the measuring point negligible.

Location fidelity is a fundamental tool for the accuracy of robotic measurement systems. Since only the actuator and not the sensor is checked, only variances of the individual robot components are taken into account. It is examined in which span the sensor head 4th is positioned on the object to be measured. It can be specified that the accuracy of the measurement location should be repeatably within a range of σ _MOT = 8mm. The self-propelled gap measuring unit 1 combines the following specified variances for measurement location accuracy. $$\begin{array}{c}{\sigma }_{G\mathrm{E.}\mathrm{S.}}={\sigma }_{K\mathrm{M.}P}+{\sigma }_{\mathrm{L.}\mathrm{B.}\mathrm{R.}}+{\sigma }_{\mathrm{M.}K\mathrm{S.}}\\ {\sigma }_{G\mathrm{E.}\mathrm{S.}}=10mm+0.3mmm+3mm=13.3mm\end{array}$$

1. 1. Das System steht auf einer definierten Position und startet die Mono-Kamera-Messung um die Position der Karosserie relativ zum Messsystem zu bestimmen. Hierbei entsteht die Varianz σ_MKS.
2. 2. Das System fährt nach abschließen des ersten Schrittes auf die erste Messposition (P1). Hierbei fällt die Varianz der fahrbaren Plattform an σ_KMP
3. 3. An der Messposition angekommen fährt der Leichtbauroboter auf den zu messenden Spalt und erzeugt so die letzte Varianz der Kette σ_LBR
4. 4. Der erste Messpunkt, auf dem der Sensorkopf steht wird als Referenz verwendet. Dabei wird der aktuelle virtuelle Spaltpunkt mit dem ursprünglichen Spaltpunkt verglichen. Durch diesen Korrekturvektor kann der Sensorkopf vom Leichtbauroboter auf die exakt gleiche Stelle aufgerichtet werden. Hier werden alle vorherigen Varianzen korrigiert und es wird der σ_Korrektur erfasst.

This variance as measurement location accuracy is not sufficient to meet the above-mentioned specification of σ _MOT = 8mm. Due to the sensor head used here 4th a relative position at the measurement location can be determined using the virtual gap points. The virtual cleavage points are fixed points in the recorded point cloud of the Sensor image. Because the sensor head 4th is not reflected in the accuracy of the measurement location, because the system only aligns it with the gap to be measured and does not perform any movement itself, it is to be regarded as an independent control element. Therefore, the data from the sensors can be used for improvement. By correcting the discount to the gap point sought, σ _GES <2mm can be assumed. Here, the currently recorded virtual gap point is compared with the position of the gap point serving as a reference. 2 shows a schematic of a measurement sequence and the variances that occur. The measurement process is divided into four steps.

1. 1. The system is in a defined position and starts the mono camera measurement to determine the position of the body relative to the measurement system. This creates the variance σ _MKS .
2. 2. After completing the first step, the system moves to the first measuring position ( P1 ). Here, the variance of the mobile platform falls on σ _KMP
3. 3. Arrived at the measuring position, the lightweight robot drives to the gap to be measured and thus generates the last variance of the chain σ _LBR
4. 4. The first measuring point on which the sensor head stands is used as a reference. The current virtual split point is compared with the original split point. With this correction vector, the sensor head can be set up by the lightweight robot in exactly the same place. All previous variances are corrected here and the σ _{correction is} recorded.

A motor vehicle is shown 6th and the self-propelled gap measuring unit 1 in a first position P1 and a second position P2 .

For the position 1 the following equation results: $${\sigma }_{G\mathrm{E.}\mathrm{S.}}={\sigma }_{\mathrm{M.}K\mathrm{S.}}+{\sigma }_{K\mathrm{M.}P}+{\sigma }_{\mathrm{L.}\mathrm{B.}\mathrm{R.}}-{\sigma }_{KOrrektur}<2mm$$

These variances are compatible with the above-mentioned specification of σ _MOT = 8mm for accuracy of the measurement location and can be verified by test series. Several series of tests were used to check the variance with which the sensor head 4th is positioned at the gap. If you combine the test series and create statistics, the following parameters are relevant. The absolute variance at the measuring location results from the individual variances in an X-direction (position to the measuring location in the direction of travel), Y-direction (height above ground) and Z-direction (distance to the measuring location). The individual variances are combined in order to calculate an amount for the actual measurement location accuracy. $${\sigma }_{\mathrm{M.}OT}=\sqrt{{\sigma }_x^2+{\mathrm{<figure-callout\; id="2"\; label="\sigma \; y"\; filenames="DE102020006436A1\_0005.png"\; state="\{\{state\}\}">\sigma </figure-callout>}}_y^{}+{\mathrm{<figure-callout\; id="2"\; label="\sigma "\; filenames="DE102020006436A1\_0005.png"\; state="\{\{state\}\}">\sigma </figure-callout>}}_2^2}$$

With the values from the test series, the following result is obtained. $${\sigma }_{\mathrm{M.}OT}=\sqrt{1.98m{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="DE102020006436A1\_0005.png"\; state="\{\{state\}\}">m</figure-callout>}}^2+{0.3}^2+1.83m{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="DE102020006436A1\_0005.png"\; state="\{\{state\}\}">m</figure-callout>}}^2}=2.71mm$$

This is the self-propelled gap measuring unit 1 in all test series around 70% more accurate than the above specification of σ _MOT = 8mm. In other words: only about a third of the tolerance that is specified as a guideline is required.

QUOTES INCLUDED IN THE DESCRIPTION

This list of the documents listed by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent literature cited

• DE 102009039812 A1 [0003]

Claims (2)

Self-propelled gap measuring unit (1), comprising a mobile platform (2), a robot (3), a sensor head (4) and a computer unit, the robot (3) being attached to the mobile platform (2) and an arm with several joints wherein the sensor head (4) is arranged at the end of the arm of the robot (3), characterized in that the sensor head (4) has two triangulation sensors, wherein a camera (5) designed as a mono camera system is also provided , wherein the camera (5) is attached to a boom which is also attached to the mobile platform (2), the mobile platform (2) having two laser scanners which are configured for cooperation according to a SLAM algorithm, the triangulation sensors are configured to work with different wavelengths, the camera (5) being configured to carry out referencing by means of edge-based tracking, with edge-based tracking Kan ten of a body recognized in an image field of the camera can be compared with edges of a stored CAD model. Use of the self-propelled gap measuring unit (1) for measuring gaps and transitions between add-on parts on a body of a motor vehicle.

Images