WO2022029808A1 - Automated predictive maintenance method of vehicles - Google Patents

Automated predictive maintenance method of vehicles Download PDF

Info

Publication number
WO2022029808A1
WO2022029808A1 PCT/IT2021/050016 IT2021050016W WO2022029808A1 WO 2022029808 A1 WO2022029808 A1 WO 2022029808A1 IT 2021050016 W IT2021050016 W IT 2021050016W WO 2022029808 A1 WO2022029808 A1 WO 2022029808A1
Authority
WO
WIPO (PCT)
Prior art keywords
vehicle
failure
exhaustion
probability
survival
Prior art date
Application number
PCT/IT2021/050016
Other languages
French (fr)
Inventor
Emanuele Pedrona
Original Assignee
Ai Parts S.R.L.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ai Parts S.R.L. filed Critical Ai Parts S.R.L.
Publication of WO2022029808A1 publication Critical patent/WO2022029808A1/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/20—Administration of product repair or maintenance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/08—Logistics, e.g. warehousing, loading or distribution; Inventory or stock management
    • G06Q10/087—Inventory or stock management, e.g. order filling, procurement or balancing against orders
    • G06Q50/40—

Definitions

  • the invention relates to an automated predictive maintenance method for vehicles, that is, a method that allows to optimize and foresee repairs or advance replacement of parts, parts or assemblies of parts and I or parts of vehicles before malfunctions or failures occur.
  • a breakage can occur unexpectedly, regardless of the design and consequent calculation of the operational life of the different pieces or parts of vehicles, for example due to particularly heavy use conditions or particularly aggressive or hostile work environment (in which a vehicle moves), or even premature wear of the parts or real "failures" occur, meaning with this expression one or more events such that the vehicle undergoes an unwanted alteration of its working conditions.
  • the invention herein illustrated and claimed here is intended to deal with overcoming the drawbacks mentioned above, thus offering a predictive maintenance method and a relatedly automated system that can deal with the occurrence of failures in a preventive manner (i.e. before they occur) and which can prepare in the shortest possible time and with the greatest possible accuracy the spare part or set of spare parts which are exactly intended to resolve that particular fault.
  • the invention also aims at devising of a method that guarantees the achievement - or otherwise stated, the learning result - of the optimal supply conditions and accuracy with respect to almost every request(s) for repair of a fault in the shortest possible “training time”, to the full advantage of both the managers/users of the vehicle faulty, of the “intervention” operators who must repair the fault, and of the logistics operators of spare parts warehouses who must provide the appropriate materials or spare parts to intervention operators.
  • the automated predictive maintenance method of vehicles according to the present invention is in the first instance initiated by a step of preparing a quantity of so-called “datasets”, which in turn relate to at least one failure of a vehicle (or as will be seen below, a breakdown or “exhaustion” of a vehicle or a part of it or of a functional group of it, as for example it can happen in the case the discharge of an electric battery, in the loss of refrigerant liquid from the circuit cooling, from the breakdown of an internal component of the engine and so on).
  • At least one dataset constitutes a set of data structured in relational form and such data are related to a fault or to a depletion of a vehicle or part of said vehicle or a functional group of said vehicle.
  • the method provides for a step of elaborating, starting from the datasets, a time period of probable breakdown or exhaustion of the vehicle (or of the vehicle part or of a vehicle functional group): this time period of probable failure or exhaustion is determined starting from an instant of initial time in which the vehicle or the vehicle part or the vehicle functional group is in a condition of maximum functional efficiency.
  • the method continues with a step of correlating to the time period of probable failure or exhaustion, of a predetermined number of objects that can be located in a warehouse or similar structure: in fact, these objects are known to technicians with the denomination of "spare parts" and therefore they are objects capable of restoring a condition of maximum functional efficiency of the vehicle or vehicle part or functional group of vehicle following the occurrence of the aforementioned breakdown or exhaustion.
  • the method at this point provides for a step of sending, by a user/requestor who will be identified later in this description, a conditional request for extraction from the warehouse of such objects or spare parts: this conditional request therefore is formulated as a function of a repair or replacement or maintenance intervention to the vehicle or vehicle part or vehicle functional group.
  • the step of sending said conditional request is performed within a time period of intervention less than the period time of probable failure or exhaustion, taking of course care that the intervention time period is calculated starting from an instant of initial time (that is, a time in which the vehicle or part of the vehicle or the functional group of vehicle is in a condition of maximum functional efficiency).
  • the step of preparing the datasets relating to at least one breakdown of a vehicle includes a sub-step of collecting usage data of a vehicle or vehicles of the same type and/or version and/or set-up, which are subsequently manipulated, in the elaborate a time period of probable failure or a probable exhaustion, through various analyzes based on artificial intelligence algorithms.
  • artificial intelligence algorithms define, based on usage data, corresponding models for predicting a probability of failure or exhaustion of the vehicle or vehicle part or vehicle functional group as a function of an operating time of the vehicle or part of the vehicle or group functional of the vehicle themselves, and in order to achieve adequate accuracy in determine such prediction, usage data may include (preferably but not limitedly to) one or more of the following categories of parameters or information:
  • the step of preparing a given quantity of dataset further includes a sub-step of determining at least an effective time period of failure or of exhaustion of the vehicle or vehicle part or vehicle functional group, and this just mentioned sub-step is performed through a statistical collection of data relating to actual failure or exhaustion events.
  • the present method can therefore determine a predetermined number of theoretical curves of a theoretical probability of failure or exhaustion of the vehicle or vehicle part or functional group of vehicle as a function of an operating time (and these theoretical curves are therefore correlated to said prediction models defined by artificial intelligence algorithms) but can remarkably determine at least one empirical- statistical curve of a probability statistics of breakdown or exhaustion of the vehicle or of the vehicle part or assembly functional vehicle as a function of an operating time of the vehicle or part of the vehicle or of the vehicle functional group themselves: this last empirical-statistical curve is therefore related to the statistical collection of data relating to real failure or exhaustion events.
  • Theoretical probabilities of failure or exhaustion generally have an increasing trend over time starting from an initial time when the vehicle or part of the vehicle or functional group of vehicle is in conditions of maximum functional efficiency towards a final time corresponding to the actual time period of failure or exhaustion or corresponding to the time period of probable failure or exhaustion: that said, the method can conveniently also comprising a step of determining one or more theoretical probabilities of survival and at least one empirical-statistical probability of survival of the vehicle or vehicle part or vehicle functional group, and such a probability of survival is inversely proportional to the corresponding probability of failure or exhaustion according to appropriate algebraic- mathematical formulations.
  • the method may also include a step of selecting and/or modifying said artificial intelligence algorithms, for example by comparison between the theoretical probabilities of failure or exhaustion and the empirical-statistic probability of failure or exhaustion (or, given the relationship of complementarity mentioned above, between the theoretical probabilities of survival and the empirical-statistical probability of survival).
  • the just introduced comparison can be useful and/or capable of determining at least one artificial intelligence algorithm or a prediction model associated with an artificial intelligence algorithm that determines a minimum or zero difference between the probabilities of theoretical failure or exhaustion and the empirical- statistical probability of failure or exhaustion and/or between the theoretical survival probabilities and empirical-statistical survival probability, and therefore can be useful to establish, even dynamically and “in self-updating”, an iterative process of refining the method itself: in this way the present method may be rendered able to "learn", for every possible variety of failure or exhaustion, the artificial intelligence algorithm or algorithms that will best simulate/reproduce/modelize the real dynamics of maturation and the occurrence of the failure or exhaustion itself.
  • the present method may further encompass a step of quantitatively and/or qualitatively visualizing an instantaneous datum of a condition of wear and/or a probability of failure and/or a probability of survival of the vehicle and/or the vehicle part and/or the vehicle functional group to a user subject (who in turn can conveniently be a user of a vehicle, a maintenance and/or repair operator of a vehicle or an operator of a spare parts warehouse).
  • conditional request submission phase can typically be after the visualization of the instantaneous data of a wear condition and/or a probability of failure and/or a probability of survival.
  • the method generates and allows the user of the vehicle to view alphanumeric data or a graphical representation (such as an arc or bar chart, or a figure with a certain variable chromatism or a combination of all possible display modes known today in the field of on-board instruments of any type of vehicle) that communicates the state of wear of a given part or of a functional group of the vehicle it is driving: this state of wear is intimately correlated with or corresponds to a probability of failure or a probability of survival, and is therefore determined by the method in accordance with the invention.
  • a graphical representation such as an arc or bar chart, or a figure with a certain variable chromatism or a combination of all possible display modes known today in the field of on-board instruments of any type of vehicle
  • the vehicle user can arbitrarily decide that the state of wear and/or the probability of failure and/or the probability of survival that is displayed is such as to require intervention in terms of "prudential" maintenance and anticipated with respect to the ordinary maintenance date, inter-correlating with the other two subjects introduced above and then planning with advance non-routine maintenance (or the method can provide that the maintenance or repair intervention can already be established in an automated manner and then communicated to the various inter-related subjects, who will only have to worry about making the intervention possible).
  • a voice message is dictated or recorded, such mecssage being related to the forwarding of the conditional request in a transmission device (and such transmission device can conveniently be a smartphone, a tablet or a computer device equipped with a sound card and a microphone and software for voice recognition);
  • the voice message is transformed into a text string and then catalogued in order to (at least) determine the vehicle and/or the failure and/or the exhaustion for which the conditional request has been submitted;
  • the text string is sent to a management unit able to manage this request conditional at least in terms of a response and/or set-up timing, of unique identification of the vehicle and unique identification of the fault or of the exhaustion;
  • a further step of making the voice message (or the text string) available to a manager can also be added: such availability may be directed to a manager of a spare parts warehouse and/or to a vehicle user and/or to a vehicle maintenance and/or repair operator, in order to establish the inter-connection between these different subjects preparing and therefore them for the material and (moreover) actually- preventive execution of the maintenance and/or repair of the vehicle.
  • the invention achieves various advantages.
  • the applications of the method according to the invention are almost infinite, since it can be adapted to various provisions or accumulations of objects (spare parts or consumable parts of any nature and/or state of aggregation, such as brake pads, seals, oils and various functional liquids and so on) of any kind, and thus responding to conditional requests of an equally variable nature with the same rapid operating and learning times and almost total accuracy.

Abstract

A predictive maintenance method comprises: - preparing datasets relating to failures; - processing, starting from datasets, a period of probable failure; - correlating spare parts to said period of probable failure; and - sending a conditional request for the extraction of spare parts, the step of preparing datasets comprising a sub-step of collecting usage data of a vehicle and being carried out through artificial intelligence, the method further comprising: - determining theoretical curves of a theoretical probability of failure related to said prediction models defined by artificial intelligence algorithms; - determining an empirical-statistical curve of a statistical probability of failure related to said statistical collection of data relating to real failure or exhaustion events; and - selecting said artificial intelligence algorithms according to a comparison between the theoretical probabilities of failure and the empirical-statistical probability of failure and/or the theoretical probabilities of survival and the empirical-statistical probability of survival.

Description

AUTOMATED PREDICTIVE MAINTENANCE METHOD OF VEHICLES
DESCRIPTION
The invention relates to an automated predictive maintenance method for vehicles, that is, a method that allows to optimize and foresee repairs or advance replacement of parts, parts or assemblies of parts and I or parts of vehicles before malfunctions or failures occur.
In various areas related to the use of vehicles, it is usually probable that certain parts of the vehicles themselves, whether they are "consumable" or designed and built to have a given operational life, must be changed, maintained or disassembled and checked: such transactions are generally expected within given time periods, and are generally included in the so-called "ordinary maintenance".
However, a breakage can occur unexpectedly, regardless of the design and consequent calculation of the operational life of the different pieces or parts of vehicles, for example due to particularly heavy use conditions or particularly aggressive or hostile work environment (in which a vehicle moves), or even premature wear of the parts or real "failures" occur, meaning with this expression one or more events such that the vehicle undergoes an unwanted alteration of its working conditions.
To deal with faults, disassembly, inspection, possible repair or replacement and re-assembly of the vehicle (or that part of the vehicle where the fault occurred) are normally required, and in order to carry out this sequence of operations a human operator must usually be required, being he/she a clerk/warehouse worker or more in general being one who has the ability to select, in a group of objects set aside or in any case kept in a "warehouse" or similar structure, one or more objects to be extracted from the warehouse itself and to be made available to whom will be physically carrying out the sequence of operations itself.
Nevertheless, the identification of the objects necessary to repair a fault can have several drawbacks, given for example by the fact that a spare parts warehouse for cars or vehicles in general must be able to be “interrogated” with sufficient precision and with technical understanding of the fault, in order to identify the right set of spare parts extractable from the warehouse itself: this process is therefore strongly subject to human errors of communication or evaluation, which result in a substantial lengthening of repair time or even the impossibility of carrying out an intervention.
Furthermore, the “on-demand” request for spare parts to repair a fault only afterwards that this has occurred often involves considerable time delays, which can be highly inconvenient as they prevent the use of the vehicle: this drawback is particularly serious where faulty vehicles are “by nature” destined to commercial or industrial use, or where they are intended for the transport of goods or passengers.
These problems are further amplified where the objects or "spare parts" to be extracted from a warehouse are of a very variable nature, and/or where possible requests' conditions are equally varied: for example, one can consider the thousands of possibilities of breakdowns of a vehicle and the corresponding requests for spare parts (single or combined with each other) and the consequent possibility of supplying/extracting parts from the warehouse spare parts, single or combinable, which can only apparently solve the faulty but which during assembly prove themselves to be unsuitable for the purpose.
In this situation, the invention herein illustrated and claimed here is intended to deal with overcoming the drawbacks mentioned above, thus offering a predictive maintenance method and a relatedly automated system that can deal with the occurrence of failures in a preventive manner (i.e. before they occur) and which can prepare in the shortest possible time and with the greatest possible accuracy the spare part or set of spare parts which are exactly intended to resolve that particular fault.
The invention also aims at devising of a method that guarantees the achievement - or otherwise stated, the learning result - of the optimal supply conditions and accuracy with respect to almost every request(s) for repair of a fault in the shortest possible “training time”, to the full advantage of both the managers/users of the vehicle faulty, of the “intervention” operators who must repair the fault, and of the logistics operators of spare parts warehouses who must provide the appropriate materials or spare parts to intervention operators.
An embodiment of the invention is therefore described, clearly by way of example and not of constraint on the field of protection and in any case subject to structural variations always included in the spirit of the invention, and that mainly comprises the following phases as described and claimed below.
More specifically, the automated predictive maintenance method of vehicles according to the present invention is in the first instance initiated by a step of preparing a quantity of so-called "datasets", which in turn relate to at least one failure of a vehicle (or as will be seen below, a breakdown or "exhaustion" of a vehicle or a part of it or of a functional group of it, as for example it can happen in the case the discharge of an electric battery, in the loss of refrigerant liquid from the circuit cooling, from the breakdown of an internal component of the engine and so on).
In general, it should be noted that in the spirit of the invention, at least one dataset constitutes a set of data structured in relational form and such data are related to a fault or to a depletion of a vehicle or part of said vehicle or a functional group of said vehicle.
Subsequently, the method provides for a step of elaborating, starting from the datasets, a time period of probable breakdown or exhaustion of the vehicle (or of the vehicle part or of a vehicle functional group): this time period of probable failure or exhaustion is determined starting from an instant of initial time in which the vehicle or the vehicle part or the vehicle functional group is in a condition of maximum functional efficiency.
The method continues with a step of correlating to the time period of probable failure or exhaustion, of a predetermined number of objects that can be located in a warehouse or similar structure: in fact, these objects are known to technicians with the denomination of "spare parts" and therefore they are objects capable of restoring a condition of maximum functional efficiency of the vehicle or vehicle part or functional group of vehicle following the occurrence of the aforementioned breakdown or exhaustion.
The method at this point provides for a step of sending, by a user/requestor who will be identified later in this description, a conditional request for extraction from the warehouse of such objects or spare parts: this conditional request therefore is formulated as a function of a repair or replacement or maintenance intervention to the vehicle or vehicle part or vehicle functional group.
Advantageously, in accordance with the present method the step of sending said conditional request is performed within a time period of intervention less than the period time of probable failure or exhaustion, taking of course care that the intervention time period is calculated starting from an instant of initial time (that is, a time in which the vehicle or part of the vehicle or the functional group of vehicle is in a condition of maximum functional efficiency).
Furthermore, according to another aspect of the invention, the step of preparing the datasets relating to at least one breakdown of a vehicle includes a sub-step of collecting usage data of a vehicle or vehicles of the same type and/or version and/or set-up, which are subsequently manipulated, in the elaborate a time period of probable failure or a probable exhaustion, through various analyzes based on artificial intelligence algorithms.
In accordance with the invention, artificial intelligence algorithms define, based on usage data, corresponding models for predicting a probability of failure or exhaustion of the vehicle or vehicle part or vehicle functional group as a function of an operating time of the vehicle or part of the vehicle or group functional of the vehicle themselves, and in order to achieve adequate accuracy in determine such prediction, usage data may include (preferably but not limitedly to) one or more of the following categories of parameters or information:
- mileage and/or time frames of operation of the vehicle or vehicles;
- identification data of the vehicle manufacturer or part of the vehicle or functional vehicle group;
- qualitative and/or quantitative indices of the production of the vehicle manufacturer or part of the vehicle or functional vehicle group; and/or - type of routes performed with the vehicle or with the vehicles; and/or
- methods of intervention on the controls of the vehicle or vehicles by the respective users; and/or
- descriptive parameters of vehicle dynamics; and/or
- environmental conditions in which the vehicle or vehicles were used.
In greater detail, the step of preparing a given quantity of dataset further includes a sub-step of determining at least an effective time period of failure or of exhaustion of the vehicle or vehicle part or vehicle functional group, and this just mentioned sub-step is performed through a statistical collection of data relating to actual failure or exhaustion events.
Having both artificial intelligence analysis of usage data and statistical collection on actual failure or exhaustion events, the present method can therefore determine a predetermined number of theoretical curves of a theoretical probability of failure or exhaustion of the vehicle or vehicle part or functional group of vehicle as a function of an operating time (and these theoretical curves are therefore correlated to said prediction models defined by artificial intelligence algorithms) but can remarkably determine at least one empirical- statistical curve of a probability statistics of breakdown or exhaustion of the vehicle or of the vehicle part or assembly functional vehicle as a function of an operating time of the vehicle or part of the vehicle or of the vehicle functional group themselves: this last empirical-statistical curve is therefore related to the statistical collection of data relating to real failure or exhaustion events.
Theoretical probabilities of failure or exhaustion (and, similarly, the probability of failure or statistical exhaustion) generally have an increasing trend over time starting from an initial time when the vehicle or part of the vehicle or functional group of vehicle is in conditions of maximum functional efficiency towards a final time corresponding to the actual time period of failure or exhaustion or corresponding to the time period of probable failure or exhaustion: that said, the method can conveniently also comprising a step of determining one or more theoretical probabilities of survival and at least one empirical-statistical probability of survival of the vehicle or vehicle part or vehicle functional group, and such a probability of survival is inversely proportional to the corresponding probability of failure or exhaustion according to appropriate algebraic- mathematical formulations.
As an example of the above, the following formulation can be considered:
[probability of survival] = 1 - [probability of failure]
In accordance with the invention and being able to dispose of both the data processed through artificial intelligence and real statistical data, the method may also include a step of selecting and/or modifying said artificial intelligence algorithms, for example by comparison between the theoretical probabilities of failure or exhaustion and the empirical-statistic probability of failure or exhaustion (or, given the relationship of complementarity mentioned above, between the theoretical probabilities of survival and the empirical-statistical probability of survival).
The just introduced comparison can be useful and/or capable of determining at least one artificial intelligence algorithm or a prediction model associated with an artificial intelligence algorithm that determines a minimum or zero difference between the probabilities of theoretical failure or exhaustion and the empirical- statistical probability of failure or exhaustion and/or between the theoretical survival probabilities and empirical-statistical survival probability, and therefore can be useful to establish, even dynamically and “in self-updating”, an iterative process of refining the method itself: in this way the present method may be rendered able to "learn", for every possible variety of failure or exhaustion, the artificial intelligence algorithm or algorithms that will best simulate/reproduce/modelize the real dynamics of maturation and the occurrence of the failure or exhaustion itself.
In even deeper detail, it can be seen that the present method may further encompass a step of quantitatively and/or qualitatively visualizing an instantaneous datum of a condition of wear and/or a probability of failure and/or a probability of survival of the vehicle and/or the vehicle part and/or the vehicle functional group to a user subject (who in turn can conveniently be a user of a vehicle, a maintenance and/or repair operator of a vehicle or an operator of a spare parts warehouse). The different possible user subjects just introduced can be mutually "interfaceable" with each other, meaning by this expression the fact that the subjects can be in mutual communication and the respective "functions" and requests can be related through the present invention: in this possible intercorrelation, the conditional request submission phase can typically be after the visualization of the instantaneous data of a wear condition and/or a probability of failure and/or a probability of survival.
To better illustrate what has just been exemplified, one may think for example of the fact that the method generates and allows the user of the vehicle to view alphanumeric data or a graphical representation (such as an arc or bar chart, or a figure with a certain variable chromatism or a combination of all possible display modes known today in the field of on-board instruments of any type of vehicle) that communicates the state of wear of a given part or of a functional group of the vehicle it is driving: this state of wear is intimately correlated with or corresponds to a probability of failure or a probability of survival, and is therefore determined by the method in accordance with the invention.
Thanks to the display made available by the invention, the vehicle user can arbitrarily decide that the state of wear and/or the probability of failure and/or the probability of survival that is displayed is such as to require intervention in terms of "prudential" maintenance and anticipated with respect to the ordinary maintenance date, inter-correlating with the other two subjects introduced above and then planning with advance non-routine maintenance (or the method can provide that the maintenance or repair intervention can already be established in an automated manner and then communicated to the various inter-related subjects, who will only have to worry about making the intervention possible).
Focusing again onto the step of forwarding the conditional request, the following sub-steps of development or implementation can be exemplified (in a non limiting embodiment of the present invention): a voice message is dictated or recorded, such mecssage being related to the forwarding of the conditional request in a transmission device (and such transmission device can conveniently be a smartphone, a tablet or a computer device equipped with a sound card and a microphone and software for voice recognition);
- the voice message is transformed into a text string and then catalogued in order to (at least) determine the vehicle and/or the failure and/or the exhaustion for which the conditional request has been submitted;
- the text string is sent to a management unit able to manage this request conditional at least in terms of a response and/or set-up timing, of unique identification of the vehicle and unique identification of the fault or of the exhaustion; and
- objects or spare parts are arranged, collected and dispatched from the warehouse to a subject who formulated said voice message.
Remarkably, in the just listed sequence of sub-steps a further step of making the voice message (or the text string) available to a manager can also be added: such availability may be directed to a manager of a spare parts warehouse and/or to a vehicle user and/or to a vehicle maintenance and/or repair operator, in order to establish the inter-connection between these different subjects preparing and therefore them for the material and (moreover) actually- preventive execution of the maintenance and/or repair of the vehicle.
In accordance with the invention and in order to make it implementable on modern telematic systems (which therefore provide for the presence and interconnection in the network of appropriate electronic devices with data storage and processing capabilities, as well as with ability to transmit and receive such data and such data processing results), one or more or all the steps of the method described up to now (and subsequently claimed), that is:
- preparing the datasets;
- processing a time period of probable failure or exhaustion;
- correlating a predetermined number of objects that can be located in a warehouse to the time period of probable failure or exhaustion; sending a conditional request;
- determining the theoretical curves of a theoretical probability of failure or exhaustion;
- determining the empirical-statistical curve of a statistical probability of failure or exhaustion;
- determining the theoretical probabilities of survival;
- determining the empirical-statistical probability of survival;
- selecting and/or modifying artificial intelligence algorithms;
- quantitatively and/or qualitatively visualizing an instant datum or data of a condition of wear and/or a probability of failure and/or a probability of survival; and
- making said voice message or text string available to a manager of a spare parts warehouse and/or to a vehicle user and/or to a vehicle maintenance and/or repair operator, are mutually connectable in at least one feedback loop and are preferably implementable through at least one “machine learning” and/or “deep learning” software.
The invention achieves various advantages.
The first of these advantages is undoubtedly the remarkable ability to increase accuracy in the supply of objects "extracted" from the warehouse, combined with the substantial zeroing of "dead time" between a fault and its repair: even more remarkably, the predictive implementation of the method according to the invention allows not only to eliminate time inefficiencies in the spare parts supply process, but even of be able to schedule preventive maintenance on the vehicle in due advance, before the fault (or failure...or breakdown) itself can occur. The applications of the method according to the invention are almost infinite, since it can be adapted to various provisions or accumulations of objects (spare parts or consumable parts of any nature and/or state of aggregation, such as brake pads, seals, oils and various functional liquids and so on) of any kind, and thus responding to conditional requests of an equally variable nature with the same rapid operating and learning times and almost total accuracy.
Moreover, numerous functional integrations of the present method into as many systems are possible, for example in terms of logistics also composed of human and/or automated/robotic hardware resources, all without affecting the basic functional capabilities of the method itself.

Claims

AUTOMATED PREDICTIVE MAINTENANCE METHOD OF VEHICLESCLAIMS
1 . automated predictive vehicle maintenance method, comprising the following steps:
- preparing a given number of datasets relating to at least one vehicle failure, at least one dataset comprising a set of relationally-structured data relating to a failure or exhaustion of a vehicle or a part of said vehicle or a functional group of said vehicle;
- processing, starting from said number of datasets, a time period of probable failure or exhaustion of said vehicle or of said part of vehicle or of said vehicle functional group, said time period of probable failure or exhaustion being determined starting from an initial time in which the vehicle or the vehicle part or the vehicle functional group is in conditions of maximum functional efficiency;
- correlating a predetermined number of objects located in a warehouse or similar structure to said time period of probable failure or exhaustion, said objects being spare parts capable of restoring a condition of maximum functional efficiency of the vehicle or the vehicle part or the vehicle functional group following the occurrence of said failure or exhaustion; and
- sending a conditional request for the extraction of said predetermined number of objects from said warehouse, said conditional request being formulated on the basis of a repair or replacement or maintenance intervention on the vehicle or vehicle part or vehicle functional group, wherein the step of sending said conditional request being performed within a time period of intervention less than said time period of probable failure or exhaustion, said time period of intervention being calculated starting from an instant of initial time in which the vehicle or the vehicle part or the vehicle functional group is in conditions of said maximum functional efficiency, and wherein the step of preparing a given number of datasets relating to at least one vehicle failure comprises a sub-step of collecting usage data of a vehicle or of multiple vehicles of the same type and/or version and/or equipment, said usage data preferably but not limitedly comprising:
- mileage and/or time frames of operation of the vehicle or vehicles;
- identification data of the vehicle manufacturer or part of the vehicle or functional vehicle group;
- qualitative and/or quantitative indices of the production of the vehicle manufacturer or part of the vehicle or functional vehicle group; and/or
- type of routes performed with the vehicle or with the vehicles; and/or
- methods of intervention on the controls of the vehicle or vehicles by the respective users; and/or
- descriptive parameters of vehicle dynamics; and/or
- environmental conditions in which the vehicle or vehicles were used, said step of processing a time period of probable failure or probable exhaustion being carried out through analyses based on artificial intelligence algorithms of said usage data, said artificial intelligence algorithms defining corresponding prediction models of a probability of vehicle failure or exhaustion or of the vehicle part or of the vehicle functional group as a function of an operating time of the vehicle or of the vehicle part or of the vehicle functional group, characterized in that it further comprises the following steps:
- determining a predetermined number of theoretical curves of a theoretical probability of failure or exhaustion of the vehicle or of the vehicle part or of vehicle functional group as a function of an operating time, said theoretical curves being related to said prediction models defined by artificial intelligence algorithms;
- determining at least an empirical-statistical curve of a statistical probability of failure or exhaustion of the vehicle or of the vehicle part or of the vehicle functional group as a function of an operating time of the vehicle or of the vehicle part or of the vehicle functional group, said empirical-statistical curve being related to said statistical collection of data relating to real failure or exhaustion events, said theoretical and statistical probability of failure or exhaustion being increasing over time starting from an initial time in which the vehicle or the vehicle part or the vehicle functional group is in conditions of maximum functional efficiency to a final time corresponding to the actual time period failure or exhaustion or corresponding to the time period of probable failure or exhaustion; and
- selecting and/or modifying said artificial intelligence algorithms according to a comparison between:
- the theoretical probabilities of failure or exhaustion and the empirical- statistical probability of failure or exhaustion; and/o
- the theoretical probabilities of survival and the empirical-statistical probability of survival, said comparison being capable of determining at least an artificial intelligence algorithm or a prediction model associated with an artificial intelligence algorithm capable of determining a minimum or zero difference between the theoretical probabilities of failure or exhaustion and the empirical probability of failure or exhaustion and/or between the theoretical probabilities of survival and the empirical-statistical probability of survival.
2. method according to claim 1 , wherein the step of preparing a given number of datasets further comprises a sub-step of determining at least an actual period of failure or exhaustion of the vehicle or the vehicle part or of the functional vehicle group, said sub step determination of the effective time period of failure or exhaustion being performed through a statistical collection of data relating to real failure or exhaustion events.
3. method according to claim 1 or 2, wherein it further comprises a step of determining one or more theoretical probabilities of survival and at least an empirical-statistical probability of survival of the vehicle or of the part of the vehicle or of the functional group of the vehicle, said survival probabilities being inversely proportional to corresponding probabilities of failure or exhaustion.
4. Method according to any one of the preceding claims, wherein it further comprises a step of quantitatively and/or qualitatively displaying an instantaneous information related to a condition of wear and/or to a probability of failure and/or to a probability of survival of the vehicle and/or of the vehicle part and/or of the vehicle functional group to a user of a vehicle or a maintenance and/or to a repair operator of a vehicle or to a manager of a spare parts warehouse that can be interfaced with the user of the vehicle and/or with the vehicle maintenance and/or with the repair operator.
5. method according to claim 4, wherein said step of sending the conditional request follows the step of displaying said instantaneous information related to a condition of wear and/or to a probability of failure and/or to a probability of survival.
6. Method according to any of the preceding claims, wherein said conditional request sending step comprises the following sub-steps:
- vocally dictating and/or recording a vocal message related to the sending of the conditional request in a transmission device, said transmission device being preferably but not limited to a smartphone, tablet or computer device equipped with a sound card and a microphone and a voice recognition software;
- transforming said voice message into a text string and cataloguing it in order to determine at least the vehicle and/or the failure and/or exhaustion for which the conditional request is sent;
- sending said text string to a management unit capable of managing this conditional request at least in terms of a response and/or set-up timing, univocal identification of the vehicle and unequivocal identification of the failure or exhaustion; and
- to order the pick-up and dispatch of items or spare parts from the warehouse to a person who has formulated said voice message.
7. method according to claim 6, wherein further it comprises a step of making said voice message or text string available to a manager of a spare parts warehouse and/or to a user of the vehicle and/or to a vehicle maintenance and/or repair operator.
8. method according to any one of the preceding claims, where one or more or all the following steps of: - preparing the datasets;
- processing a time period of probable failure or exhaustion;
- correlating the time period of probable failure or exhaustion to a predetermined number of objects located in a warehouse;
- sending a conditional request;
- determining the theoretical curves of a theoretical probability of failure or exhaustion;
- determining the empirical-statistical curve of a statistical probability of failure or exhaustion;
- determining the theoretical probabilities of survival;
- determining the empirical-statistical probability of survival;
- selecting and/or modifying the artificial intelligence algorithms;
- quantitatively and/or qualitatively displaying the information related to a wear condition and/or to a probability of failure and/or to a probability of survival; and
- making said voice message or text string available to a manager of a spare parts warehouse and/or to a vehicle user and/or to a vehicle maintenance and/or repair operator, are mutually connectable in at least one feedback loop and are preferably implementable through at least one machine learning and/or deep learning software.
PCT/IT2021/050016 2020-08-07 2021-01-19 Automated predictive maintenance method of vehicles WO2022029808A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
IT102020000019822A IT202000019822A1 (en) 2020-08-07 2020-08-07 AUTOMATED PREDICTIVE MAINTENANCE METHOD OF VEHICLES
IT102020000019822 2020-08-07

Publications (1)

Publication Number Publication Date
WO2022029808A1 true WO2022029808A1 (en) 2022-02-10

Family

ID=72886083

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IT2021/050016 WO2022029808A1 (en) 2020-08-07 2021-01-19 Automated predictive maintenance method of vehicles

Country Status (2)

Country Link
IT (1) IT202000019822A1 (en)
WO (1) WO2022029808A1 (en)

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019213177A1 (en) * 2018-04-30 2019-11-07 Ramaci Jonathan E Vehicle telematic assistive apparatus and system

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019213177A1 (en) * 2018-04-30 2019-11-07 Ramaci Jonathan E Vehicle telematic assistive apparatus and system

Also Published As

Publication number Publication date
IT202000019822A1 (en) 2022-02-07

Similar Documents

Publication Publication Date Title
US10410135B2 (en) Systems and/or methods for dynamic anomaly detection in machine sensor data
US20090157455A1 (en) Instruction system and method for equipment problem solving
Mori et al. Development of remote monitoring and maintenance system for machine tools
Brundage et al. Where do we start? Guidance for technology implementation in maintenance management for manufacturing
US8065334B2 (en) Warranty insight solution framework system and method
US9189760B2 (en) System, method and computer readable medium for using performance indicators and predictive analysis for setting manufacturing equipment parameters
AU2023274062A1 (en) System for analyzing machine data
CN104238491B (en) Information platform for the data processing based on industrial automation stream
JP2007199795A (en) Alarm information processing apparatus and alarm information processing method
US20120116827A1 (en) Plant analyzing system
Lewis et al. Maintenance of mobile mine equipment in the information age
CN110967629B (en) Fault diagnosis system and method for electric drive device
KR20130065800A (en) Real-time collaborated enterprise asset management system based on condition-based maintenance and method thereof
CN111427330A (en) Equipment maintenance data-based equipment fault mode and rule analysis method
US20170038281A1 (en) Method of predicting life of component of machine
US11544250B2 (en) Fault finding support system and method
WO2022029808A1 (en) Automated predictive maintenance method of vehicles
US20050027487A1 (en) Product defect analysis and resolution system
KR20230098538A (en) System and Method of Hybrid Reliability Centered Maintenance for Power Generation Facilities
Sharp et al. Selecting Optimal Data for Creating Informed Maintenance Decisions in a Manufacturing Environment
Diaz-Elsayed et al. Asset Condition Management: A Framework for Smart, Health-Ready Manufacturing Systems
US11645631B2 (en) Method and system for automatic maintenance of a machine
JP2022132848A (en) Operation monitoring device and operation monitoring method
JP2003345424A (en) Apparatus maintenance management system
KR20200121082A (en) Simulation system of repair prediction for equipment asset management and the method thereof

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21705281

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21705281

Country of ref document: EP

Kind code of ref document: A1