RaspberryPiVolvoLogger
| Title | RaspberryPiVolvoLogger |
|---|---|
| Summary | RaspberryPi-based solution for logging CAN data on Volvo trucks |
| Keywords | GNU/Linux, CAN interface, Data Mining, Knowledge Representation |
| TimeFrame | Spring 2015 |
| References | http://www.raspberrypi.org/ |
| Prerequisites | Basic Linux knowledge, programming competence, possibly electronics experience,
Cooperating Intelligent Systems or equivalent basic Artificial Intelligence course |
| Author | |
| Supervisor | Slawomir Nowaczyk, Rune Prytz |
| Level | Master |
| Status | Open |
In the ReDi2Service project we are working together with Volvo Technology in Goteborg on collecting on-board data from a fleet of buses and comparing individual vehicles against rest of the group to detect faults and component wear.
In that project we are using a specialised hardware and software solution, but we are interested in exploring possibilities of using products such as Raspberry Pi http://www.raspberrypi.org/ in order to lower the costs and increase the flexibility.
Therefore, the first task in the project would be to evaluate possible solutions for connecting Raspberry Pi to the CAN networks available in the vehicle, namely http://en.wikipedia.org/wiki/J1939 and http://en.wikipedia.org/wiki/J1587
Once this connection is realised, there is significant flexibility in the project and next steps depend on students' interests. Some selection of directions mentioned below, as well as other ideas, will form the main part of the thesis.
- A flexible data logging system, allowing high-level specification of data to be collected, as well as conditions under which is should be done. This could also include high-level and (partially) learned usage and mission classification parameters.
- A communication model which allows efficient transmission of collected data to the back office, based on different assumptions concerning network architecture, for example with fixed number of base stations and using other encountered vehicles as relays.
- On-board processing of collected data, especially generation of different forms of data aggregation and compression models (e.g. histograms, linear relations, etc), both defined by user and learned using supervised and unsupervised techniques.
- An algorithms for distributed comparison of collected data across the fleet, without the need of transmitting it all to the central server, but taking advantage of the dynamic network topology.
- Targeted diagnostics and monitoring, where the definition of data being collected adapts to the current needs of individual vehicle or the fleet. The focus can shift towards components or subsystems that require more attention, based on different criteria (both external conditions as well as internal component state).
- Modularisation is crucial for flexible data logging system, facilitating reuse, via standardised communication modules, abstracting device specific functions -- thus allowing it to be used in multiple domains, for example both in automotive and health care/home monitoring setting. One task in the project could be to integrate data received over CAN interface with PostgreSQL DBMS, implementing data aggregation and compression models within the database.
