Researchers create computing fog for driverless cars

Team puts scalable, reliable fog computing systems at the roadside
25th October 2018

Researchers from across the region have been demonstrating technology developments that are vital for the next generation of driverless cars.

The £5m FLOURISH driverless car research and development project, one of several in the region, is half way through its three year investigation into the future deployment of connected and autonomous vehicles (CAVs) on the UK road network.

“At a technical level, FLOURISH supports this aspiration by investigating how data can be used to optimise regional transport networks and how to ensure secure communications between autonomous vehicles and the surrounding transport infrastructure,” said Tracey Poole, FLOURISH Project Manager.

Under the trial conditions, the maximum distance for the successful communication of messages was 472 metres, with some non-line-of-sight coverage achieved. Transmission reliability, measured in terms of the awareness horizon, showed a rapid drop-off in the delivery of reliable communications beyond 120 metres. 95% of all messages were received within 778 milliseconds. The data gathered has provided a benchmark against which the performance of a wide range of Vehicle-to-Everything (V2X) services can be evaluated.

As part of the project, the team at the University of Bristol is developing technology that can handle different types of wireless link such as the 802.11p variant of WiFi and 60GHz millimetre wave (from Blu Wireless Technology in Bristol) through a roadside access point, routing data to secure, scalable roadside computing systems.

This architecture allows the system to handle different layers of data depending on the importance of the data, the amount and the latency required, but needs sophisticated networking software technology. This ‘fog layer’ is scalable and transparent to the system providers and can also have redundant nodes to allow reliable 24/7 operation, says the team, which includes Dr Andre Tassi (above, left), Dr Ioannis Mavromatis and Dr Robert Piechocki, principal investigator (right).

The project is also looking at human factors. The early findings from user research suggest that the trust of older adults in driverless cars is optimised when they have the option of receiving journey management information from the vehicle in both an audio and visual form. These findings will help inform future HMI designs and how the cars communicate with each other and roadside infrastructure.

The driverless car trials are conducted in Bristol and South Gloucestershire. The first car trial tested a connected vehicle network and associated technologies in a complex urban environment, testing message delivery, the distance over which messages can be successfully sent (depending on the local environment) and the time between the sending and receipt of messages. It also tested success against the performance level of the transceivers, to develop an understanding of the impact of the communication equipment’s performance on the functionality of the communication service.

The project brings together researchers from Airbus, Aimsun,  AXA UK, Bristol City Council, Cardiff University, Designability, Dynniq, React AI, South Gloucestershire Council, Transport Systems Catapult, Traverse, University of Bristol and University of the West of England, with support from the Bristol Robotics Laboratory (BRL).

“Future simulator and pod trials will focus on gaining further insights into what older adults need from CAVs, looking at additional HMI features, including voice recognition,” said Poole. “Future car trials will build on the previous trials to further test the CAV communication network, cooperative services and associated technologies.”

The details of the mid-project report are at www.flourishmobility.com