Big Automotive Data Analytics (BADA) main study phase 1
Reference number | |
Coordinator | Volvo Personvagnar Aktiebolag - Volvo Personvagnar AB |
Funding from Vinnova | SEK 15 000 000 |
Project duration | March 2015 - June 2018 |
Status | Completed |
End-of-project report | 2015-00677.pdf(pdf, 459 kB) (In Swedish) |
Important results from the project
Industry-wide collaboration between industry and research institutes, but also between information and communication technology (ICT) actors, domain actors, data researchers, computer scientists and computational engineers. Collaboration between the parties AB Volvo, Scania, Volvo Cars, Trafikveket and RISE to jointly show a case where vehicles from different manufacturers talk to each other and exchange information regarding activated warning lights. These messages are also captured by RISE for compilation and analysis.
Expected long term effects
In the work with the Hazard Warning user case, BADA developed a platform for analyzing streaming data based on Apache Kafka, and also demonstrated how the use of warning flashes can be classified in real time. (see also http://bada.sics.se/?page_id=168) The work of the user cases demonstrated how state-of-the-art algorithms can be used on project prototype problems (e.g., Exploration of Accident Data, Queue Detection) on the data made available. (See also http://bada.sics.se/?page_id=178 and http://bada.sics.se/?page_id=41
Approach and implementation
BADA use HopsWorks environment where BADA data was explored in Spark and run at SICS ICE Datacenter. A tutorial is available at www.sics.se/~ahmad/traffic-flow-analysis.html to run these algorithms. BADA analysis platform (Spark, HopsWorks) was used in Hazard Warnings and in a traffic flow analysis demonstrator www.sics.se/~ahmad/traffic-flow-analysis.html). BADA´s analysis platform (Spark, HopsWorks) was used in Hazard Warning´s user case as well as in a traffic flow demonstration (https://www.sics.se/~ahmad/traffic-flow-analysis.html).