ReDi2Service Remote Diagnostic Tools and Services Niclas Karlsson, PM Volvo Group Trucks Technology & Halmstad University
Agenda Why ReDi2Service? Overview of the concept Subprojects ToDo list Publications
What s the motivation? Increase Uptime Predict faults before they happen and turn unplanned stops into planned ones Adapt service plans to actual service needs of individual vehicles Minimize over-maintenance time at workshops New Services Develop new services to increase customer value Create new business cases
Consider a fleet of city buses, buses that drive around in a city Buses are driving under similar weather and load conditions (number of passengers, hills, etc.). Suppose every bus is equipped with equipment that constantly listens to the sensor and actuator data travelling around on the internal data network.
This equipment captures and analyzes relationships between the signals on the internal network
Every now and then the buses report to a central server, via a telematics gateway, those relationships that look interesting.
Every now and then the buses report to a central server, via a telematics gateway, those relationships that look interesting. Deviations are detected and flagged for repair / service. Diagnosis can be done onboard and maintenance needs can be predicted. These are collected and compared between vehicles and to fault signatures.
Technology development Service development
Data Collection Developed an On-board data logger (Volvo VACT system) Logging data since September 2011 Collecting data on a bus fleet Data from trucks Logging ~100 signals at 1 Hz ~250GB of data Access to maintenance records and driver comments Similar data is being collected or will be collected soon in other projects (e.g. InnoMerge)
c Individual Signal Analysis Many more deviations than we originally expected Comparing individuals against the fleet Still unknown how serious these issues are Tp 0.7 0.6 0.5 0.4 0.3 0.2 N1 N2 N3 N4 A33 A50 G50 G100 CAC38 CAC316 CAC332 Ex30 Ex35 Ex40 0.1 0 0.2 0.4 0.6 b 0.8 1 0.08 0.07 0.06 a 0.05 0.04 0.03 0.02
Relationships between signals One strong relation is related to wheel speed sensors. There are clear deviations in these that seem predictive of certain faults. Work on relationships is ongoing.
Current Work Comparing single vehicle at different times There are clear trends, but we are still working on ways to factor out external conditions (without using the fleet) Relating deviations in the data to Vehicle Service Records database records Difficult to get reliable information about faults There are correlations in some, but not all, cases
Services enabled by ReDi2Service Maintenance planning Workshop Volvo internal Stand-alone service for fleet customers Diagnostics tools Planning (tools, parts, instructions) ReDi2Service On-board services Driver aid Parameter settings Purchasing decisions Management Follow up information Product development Operations information Driver behavior Traffic information
Service Development Summary What has been done? Stakeholder, external, internal and risk analysis are done. Cocreation of services with customers is initialised. Results so far Papers Business ideas/prospects Completed study on one certain fleet operator (to map and improve their maintenance process) Loads of interview material to analyze
Work in progress Service concept development in co-operation with customers Design of demonstrator to illustrate service concepts and business opportunities.
Remaining work Secure signal deviations that show a real problem Identify more relevant signal deviations Controlled test Fault injection experiment Build a demonstrator Finalize development of new services Transfer results and knowledge to stakeholders
Publications part I(II) Journal M. Svensson, S. Byttner and T. Rögnvaldsson, Vehicle Diagnostics Method by Anomaly Detection and Fault Identification Software, To app SAE Transactions (2009) [also 2009 SAE International Congress, SAE Technical Report SAE 2009-01-1028] S. Byttner, T. Rögnvaldsson and M. Svensson, Automatic fault detection with self-organizing models and vehicle telematics, in revision for Engineering Applications in Artificial Intelligence, 2009. Paper A. Mosallam, S. Byttner, M. Svensson, T. Rögnvaldsson, Nonlinear Relation Mining for Maintenance Prediction to be presented at IEEE Ae Conference AIAA, Montana, USA, Mars 5-12, 2011 G. Vachkov, S. Byttner, M. Svensson, Incremental Classification of Process Data for Anomaly Detection Based on Similarity Analysis to be presented at IEEE Symposium Series on Computational Intelligence, Paris, A 10-15 2011 S. Byttner, M. Svensson and T. Rögnvaldsson, Finding the Odd-One-Out in Fleets of Mechatronic Systems using Embedded Intellient Agent AAAI Spring Symposium, Stanford, CA, March 22-24 2010. M. Svensson, M. Forsberg, S. Byttner and T. Rögnvaldsson, Deviation Detection by Self-organized On-line Models Simulated on a Feed-bac Controlled DC-motor, IASTED conference on Intelligent Systems and Control, Cambridge, MA, Nov. 2-4, 2009. S. Byttner, T. Rögnvaldsson, M. Svensson, G. Bitar and W. Chominsky, Networked vehicles for automated fault detection, 2009 Internationa Symposium on Circuits and Systems, May 24-27, Taipei, Taiwan (2009) T. Rögnvaldsson, G. Panholzer, S. Byttner and M. Svensson, A self-organized approach for unsupervised fault detection in multiple systems Proc. ICPR2008 (International Conference on Pattern Recognition), Tampa, Florida, Dec. 8-11 (2008) M. Svensson, S. Byttner, T. Rögnvaldsson, Self-organizing Maps for Automatic Fault Detection in a Vehicle Cooling System, Proc. IEEE International Conference on Intelligent Systems, Varna, Bulgaria (2008). (2nd place for best student paper) S. Byttner, M. Svensson and T. Rögnvaldsson, Self-organized modeling for vehicle fleet based fault detection, SAE Technical Report 2008-1297, SAE World Congress 2008, (2008) S. Byttner, M. Svensson and T. Rögnvaldsson, Modeling for vehicle fleet remote diagnostics, SAE Technical Report SAE 2007-01-4154 (20 Patent: Remote diagnosis modelling, PCT filed May 14, 2007 (P21673 WO).
Publications part II(II) Chowdhury, S., Akram, A. and Åkesson, M., (2012). E-maintenance as an Emerging Customer Value Generating IT-enabled Resource, Accepted for The Mediterranean Conference on Information Systems (MCIS), 8-10 September 2012, Portugal. Chowdhury, S., (2012). Co-creation of Innovative Digital Services, Submitted for The 35th Information Systems Research Conference in Scandinavia (IRIS). 17-20 August 2012, Sweden. Akram, A., (2011). Towards Servitization in the Age of Digital Innovation A Case from Vehicle Industry. Submitted for The 35th Information Systems Research Conference in Scandinavia (IRIS). 17-20 August 2012, Sweden. Chowdhury, S. and Åkesson, M., 2011. A proposed framework for identifying the logic of digital services. Accepted for 15 th Pacific Asia Conference on Information Systems. 7-11 July, 2011 Akram, A. and Åkesson, M., 2011. Value network transformation by digital service innovation in vehicle industry. 7-11 July, 2011 Chowdhury, S. and Akram, A., 2011. E-maintenance: Challenges and Opportunities. Accepted for 34 th Information Systems Research Seminar in Scandinavia. 16-19 August, 2011 Akram, A. and Åkesson, M., 2011. A research agenda to study how digital service innovation transform value network. Accepted for 34 th Information Systems Research Seminar in Scandinavia. 16-19 August, 2011 Rögnvaldsson, T., Svensson, M. and Akram, A., 2010. Uptime in the commercial vehicle business [white paper]
Thank you!