Introduction

COVESA is actively investigating essential standards and solutions needed to enable a data-oriented, connected vehicle architecture. Connected cars will become the predominant form of individual automotive transportation. A number of actors are now working at breaking down the barriers to adoption for mobility services based on automotive data and at helping drivers, communities, mobility services, aftermarket and repair services and automotive manufacturers extract the most value from that data. But this has already resulted in a somewhat fragmented ecosystem where different actors are using different solutions to access the data of connected cars.

Also it needs to be considered, that different use cases require different ways of accessing data. While training an AI model on data will not require live streamed data, the inference of the model may work on a data stream directly from a car or a fleet.

Last but not least, regulation is changing. The upcoming EU data act require the access to raw data for the data producer in a way that he can make it accessible to third parties. There is no solution yet that solves that problem. It might be noted here, people may own more than one car and it seems inappropriate to believe that customers are willing to join multiple OEM driven ecosystems.

Framing question

How can COVESA create an ecosystem that allows car manufacturers, car owners (drivers), 3rd party service providers and others to offer, monetize and consume services in a:

• secure,
• reliable,
• compliant, and
• managed

environment. To support business of scale, members of the ecosystem are required to onboard through self-services.

Insights

When collecting data from cars things get easily complicated. Cars are owned, driven by different individuals but the data is personally relatable to the owner of the car. So the use cases need to consider all possible permutations of the actors in the system, related to their produced data. The following figure displays the relations between the parties. This figure can be extended in the future to reflect more possible use cases.

Furthermore for a professionla environment of exchanging data you need to consider the following as well:

• Services are based on exchange of data (historical, actual and near realtime).
• exchange of personal identifiable information (PII) is governed by user rights based on laws.
• this implies contracts (usually referred as consent) between the user and the data controller and contracts between the controller and the processor
• data is distributed in different environments (OEM cloud space, 3rd parties ...)
• data is available in different formats (preferably VSS)
• data is protected by several access mechanisms
• near realtime data access must be optimized to preserve bandwith
• data access must be governed by SLA's

Needless to mention that all data needs to be secured by state of the art security measures and access is only granted to authenticated and trusted individuals.

Product vision

Putting the above mention knowledge in a solution COVESA should aim for providing:

A scaling and managed environment with:

• a unified access to cross OEM and service provider data,
• a well defined SLA and legally compliant,
• a proper separation of safety relevant and regulated data,
• straight access to a number of data points,
• optimized data consumption on transport side,
• upload to any data store,

protected by

• access right management,
• rate limitation,
• a paywall, and
• state of the art security measures.

Outcomes

• The OEM's and other data owning entities get the peace of mind by having access managed according to law.
• Individual contract negotiations are not required and data can be monetized more easily.
• Accessibility increases attractivity, usefulness and visibility to third party developers, and enables them to experiment with their business ideas with very low entry barriers.

To enable scale, all must be self-services. No human interaction is required to onboard car manufacturers users, cars, new services.

Two options are possible and needs to be evaluated.

1. Direct data access after onboarding (bypassing the OEM cloud) or
2. indirect access via the OEM cloud.

For different development phases, not only car data can be utilized but also recorded data in a data lake (served by the playgroud) or synthetic data.

Project proposal

Endava would like to initiate and lead a project to

• collected and considered requirements
• document of all considered use case
• define and document this ecosystem,
• implement a prototype to demonstrate the capabilities.

Our suggestion is to base this project on already existing work done prevoiusly for COVESA, see references.

References

Legal references

1. EU comission

2. ISO20078