James Long, head of technical consulting for the Smart Mobility Living Lab (SMLL), recently contributed to CIHT’s monthly masterclass on Planning for an Autonomous Vehicle Future. Among other issues, he looked at testing autonomous vehicles in real-world scenarios. We asked him the key questions.

What does the Smart Mobility Living Lab do to help the development of autonomous vehicles?

Smart Mobility Living Lab is an urban, real-world testbed, established as part of a £20m programme jointly funded by government and industry. The consortium has established a real-world test facility, because we recognised that the most critical part of any testing and trialling is exposing autonomous mobility technologies and services to real people, in real places, in the safe way.

We validate capabilities, exposing them to end users and understanding how they perform out in the real world. Because we're a public environment, we can't put a fence around where we're working, so instead we monitor the performance of new technologies and services. Anything that we can control, we do, but anything we can't, we measure. We have focused on instrumenting a network so we can understand the performance of different solutions and their impact on people and place. We also enable organisations to integrate their solutions into our network to save them from investing in a full end-to-end system before being in a position to validate their capabilities.

Can you tell us about any autonomous vehicle projects that you’re currently working on?

We're working on a project called ServCity. The focus has been on how we enable connected and automated vehicles (CAVs) to safely navigate and engage in complex environments like urban spaces. It looks at everything from the user perspective, in terms of how different user groups might interact with services – and the way that you might design for that, from a systems perspective. There's been a focus on scalability, looking at different types of junctions and road environments where users might be showing a demand for services. It also looks at the perception of solutions and how that changes in urban settings, and then how to deploy vehicles into those complex urban environments. A major component of this project has included using SMLL’s roadside infrastructure as an off-board sensor to help the CAV understand conditions on the road ahead, and to inform its decision making.

How are safety risks assessed when evaluating projects like this?

We operate under the DfT’s code of practice for automated vehicle trialling. A central element is having a safety driver in control of the vehicle at all times, and when we’re operating in complex environments with complex systems, this is underpinned by a very strong safety case process. One of the topics we have focused on for ServCity is the role of infrastructure, trying to understand how we accurately identify different objects, and the latencies associated with transmitting data to vehicles. This contributes to appraising the performance of those systems. Safety is an underlying component of this. Another element is looking at concepts around in-use monitoring and understanding how the systems work once they're deployed.

How are safety risks or challenges mitigated?

We conduct comprehensive safety cases, looking at the systems level and the operational risks. We do operational safety cases that consider hazards and risks in the environment. As part of that we look at different routes, different environments, different times of day – all the components that represent the operational design domain.

What's delaying or preventing scalability when it comes to connected and automotive mobility?

One of the topics that keeps coming up is who will pay? If there's a need to change infrastructure or data provisions, for example, there's a question mark over who is investing – and for what reason? For me, the question is not who, but why. If we can answer the question of why we need certain support for vehicles, where might that be beneficial, what are the use cases and needs of that support – from the user perspective, but also from the system perspective – then we can start to understand the performance level for that data, and the case for investment.

James Long was in conversation with Craig Thomas

Missed the masterclass featuring James Long on Planning for an Autonomous Vehicle Future? The next CIHT webinar is ‘NEC Contracts – how does it work and key highlights’ on 28 September at 8am. Free to all CIHT members.

Missed the masterclass featuring James Long on Planning for an Autonomous Vehicle Future? The next CIHT webinar is ‘NEC Contracts – how does it work and key highlights’ on 28 September at 8am. Free to all CIHT members.