Submitted by Rachel Gardner on Thu, 03/09/2020 - 11:03
Dr Amanda Prorok, an Assistant Professor and University Lecturer in Cyber-Physical Systems here, has today been awarded a coveted European Research Council Starting Grant.
'Robot trainer' Dr Prorok will be using the award to fund a futuristic – and highly ambitious study – into the transport environment of the future. She will look for ways simultaneously to optimise both the collective mobility of robots, when they are working in a group, and the environment in which they are operating.
This, she says, will be a first as the research fields of robotics and spatial design are not currently linked. "I'm really hoping to create a new interdisciplinary research community that considers the environment, and robotic mobility and robotic path-planning, jointly," she explains.
This grant is one of two ERC Starting Grants awarded today to researchers in this department. Dr Alice Hutchings is also the recipient of a Starting Grant.
ERC Starting Grants are aimed at talented early-career scientists who have completed their PhD within the last seven years, who have already produced excellent supervised work and show potential to be a research leader. This year, the ERC awarded funding worth in total €677 million, to help these early-career scientists and scholars to build their own teams and conduct pioneering research across all disciplines. The grants are part of the EU’s Research and Innovation programme, Horizon 2020.
The grants carry considerable prestige. They are awarded annually via a very competitive application process; this year, for example, just 436 of the 3,272 applications were successful – a success rate of 13.3%.
Dr Prorok's current research focuses on methods to encourage autonomous agents – such as robots and driverless cars – to collaborate and cooperate when they are working together in shared spaces.
This includes ways of helping robots to negotiate the cluttered space of a warehouse, or driverless cars to collaborate to avoid traffic jams and collisions when driving on a road partially blocked by a stationery vehicle. But it could also apply to other sorts of environments where artificially intelligent agents are moving around within a space.
That could include robot workers on farms, collecting produce from the fields and putting it into containers, or robots moving cargo around an airport from a warehouse to a plane.
But she says that if we are going to make further progress in this area, we need to stop trying to shoehorn the robot mobility systems of the future into the transport environments of the past.
"Instead of considering only the robots themselves, we should also factor in the environment itself as a variable."
And that is what she is aiming to do in her ERC-funded project, Scalable Co-optimization of Collective Robotic Mobility and the Artificial Environment.
She points out that traditional approaches to the design of mobile robot systems consider the environment – a warehouse, for example, or a road – to be a fixed constraint.
In these environments, existing structures and obstacles must be navigated, and the robots have to negotiate for right-of-way with others. And even in cooperative mobile robot systems, these constraints can lead to bottlenecks.
Up until now, structures like road junctions, roundabouts and lanes have been used to help ease these bottlenecks, improve navigation and remove driver conflicts.
But this is an old-fashioned approach, Dr Prorok argues. "As we progress with automated, roboticised mobility systems," she says, "we must jointly re-evaluate the shape, form, and function of the environments that we operate in."
So instead of trying to fit driverless cars into a road network designed for human drivers, when we are thinking about designing the cities of the future, we should be ripping up that blueprint and starting from scratch.
This, she admits, "is completely new and relatively futuristic. But we do need to think about the design of the cities of the future. Are we going to build them for the ‘current’ future – where we have people-driven cars – or are we going to build them for the 'future' future?
"If so, we need to start understanding what that looks like, so that we can design these cities properly, right from the beginning. With this ERC-funded research project – which I’m really excited about - I hope to create a new research community that thinks about this in a joint manner."
Dr Prorok is keen to start work on her research. And as she does so, she and the research staff she recruits for this work, will be keeping their minds as open as possible.
"I want to focus on the fundamentals of robotic mobility and structures and space, so that the insights we gain are as general as possible," she says.
"Then once we have a new foundational understanding of what environmental structures lead to which kinds of behaviours among connected autonomous agents, then we could translate that into specific applications."
