Data analytics in different use cases

At the 16th International CTI-Conference “Automotive Diagnostics” the participants had the chance to attend a workshop hosted by Robert Bosch GmbH. The attendees discussed in two different groups about Data analytics in different diagnostic use cases. They focused on the topics “Benefit & Value”, “Improvement & Changes” and “Regulations”.

Different questions were asked and discussed:

Benefit & Value

How could we (OEM, supplier, end user, …) benefit form the value in providing the data?

  • improve the quality Historical Data
  • avoid breakdowns
  • improve the quality of products
  • more transparency to customer
  • identify correlations
  • improve the quality of cars
  • improve the quality of onboard diagnostic data
  • avoid mistakes
  • improve efficency
  • improve developing process
  • create fun for end user
  • data for second owner of a car
  • driver profiles for fleets e.g. how the product is used
  • more comfort for end user
  • improve updates

How do you judge the value of a data analytics solution?

  • time and cost to fix a problem
  • downtime for CV
  • availability of mobility for mobility provider
  • reduction of warranty costs
  • support of quality engineers
  • availability of the right data to the different users
  • better decisions
  • new information extracted of the data
  • to judge if models/ assumptions are correct
  • correlations to events
  • safe costs
  • return of investments

Improvement & Changes

Which use cases are used to make decissions?

  • most efficient order of executing diagnostic steps
  • no future found
  • technical & business
  • business: visulisation, more up-to-date – automatisation
  • sales recomendations technical diagnosis systems
  • quality error identification
  • analyse data: large variety, sending behaviour of cars – planning of BE systems

Where do we see the biggest needs to improve data analytics?

  • infrastructure/ data collection
  •  Algorithmus/ Analytics
  • user interfacts
  • standardisation of data formats & data access
  • interpretable models
  • secure access to high quality data
  • competition of services
  • data needed to answer the use cases – not yet completly aviable
  • data validation
  • confidence estimation
  • suppliers: Adaption of systems
  • needs: date from realt usage, collegt right data
  • today: manually steps – automatisation – powerful tools reg.

How could future mobility (shuttle, autonomous driving, …) impact data analytics in diagnostics?

  • more homogenious diagnostic data – more different to analysis
  • on board preprocessing
  • usage pattern of vehicle
  • frequent updates
  • continue learning with data
  • consens from user reg. data collection
  • reduce amount of data to get valid result
  • upcomming governmental regulations

Regulations

How could governmental regulations impact data analytics?

  • emission based control & monitoring on-line monitoring
  • IDS & security regulation, e.g. V2V, V2X, …
  • RMI (EU), right to repair (US)
  • security & privacy
  • different country = different regulation
  • access to data in vehicle is not regulated
  • cybersecurity
  • Right to repair
  • GDPR
  • warranty legislations
  • accident data recorder
  • new types approved
  • NEVADA neutral server
  • tolling + Country border

How to judge the existing SAE norm to exchange data and maintenance information?

  • no knowledge
  • C-IST are EsVec but usage are open?
  • no chance of a standard?
  • no experience, no information
  • ExVec or alternative
  • ISO/ CEN Standard on RMI ISO 18541
  • example: C-IST standartisation