Illustration of dashboard control in an auto piloted simulation smart car

Digi CAV

Funder

Innovate UK

Value to Coventry University

£105,000

Project team

Dr Rahat Iqbal

Partners

HORIBA MIRA

Duration of project

01/12/2018 - 31/08/2020

Centre for Connected Autonomous Automotive Research logo

MIRA logo

Project overview

The validation of high-level autonomy features requires large amounts of test data, which conventionally is achieved by accumulating miles on the road and dedicated proving grounds. A conventional new car model costs between $1-6 billion to take through from design to commercialisation, and as much as half of this can be validation costs. For highly automated vehicles (SAE Level 4+) vehicles currently in development, the validation costs are likely to be significantly higher, which will slow the introduction and take-up through increased costs.

The aim of this project is to de-risk the technical approach to build a simulation platform that would significantly reduce these validation costs.

Project objectives

Investigating the feasibility of building a connected autonomous vehicles (CAV) simulation platform that enables plugging in external heterogeneous components such as electronic control units (ECU), autonomous driving modules, simulation software, sensor data and algorithms. Level 4 autonomy capability will be demonstrated through a Hardware-in-the-Loop (HiL) simulation solution.

Impact statement

A key feature of this platform is that it will be agnostic to the source of the component, having compatible functionality without forcing IP owners to disclose their protected methods and algorithms.

This feasibility project will de-risk the technical approach and form the foundation for a larger, follow-on project which will bring on board potential customers and IP owners to ensure the final capability is applicable throughout the supply chain.

Outputs

The initial tests were performed using Interactive Coventry’s Advanced Driving Assistance Systems (ADAS) which demonstrated the effectiveness of the project through hardware in the loop and software in the loop without revealing the product IP. In addition a number of fault injection techniques have been developed to test the resilience of the system to changing weather conditions and potential hardware faults. An interface to facilitate the exchange of data between the simulated environment and IC’s ADAS has been developed and integrated with the solution.