Scientists are developing situation-aware mobile robotic systems for use in applications such as transport, logistics, space, defence, agriculture and infrastructure management.
Led by Prof Paul Newman of Oxford’s Department of Engineering Science, the programme – which has been awarded a £5m EPSRC grant – will include numerous partners including Ricardo, Network Rail, the Fraunhofer Institute, Amey, the UK Space Agency, BP Global and Nissan.
The project’s aim is to create the world’s leading research programme in mobile autonomy. To do this, it will have to overcome some of the fundamental technical issues which have so far prevented the large scale adoption of mobile robotics by industry and society, such as a need for them to be cheap, work cooperatively with people in large, complex and time-changing environments and do so for long periods of time, all while remaining safe and trusted.
Using the mathematics of probability and estimation, computers in robots will be enabled to interpret data from sensors like cameras, radars and lasers in order to form a map of their surroundings.
Using this map, the robotic systems will be able to localise themselves, determining their own position in their environment. They will then be able to identify the objects around them and match live imagery with a preexisting database to figure out where they are, where to go, and what to avoid, taking into account changes such as reduced or increased lighting. However, the group also hopes to reduce reliance on expensive sensors by engineering technologies that lower the cost of autonomy and enable cheaper sensors to be used.
‘Surveying is an important part of our work,’ explained Prof Newman. ‘Many of our algorithms rely on detailed and accurate surveys of the environment to function. The NABU (a data collection unit and map generation device) generates detailed maps so that our robots can subsequently navigate using them. However, there are other applications for this technology which may become marketable in the near future.’
The research programme will contain a number of separate Flagship projects covering different applications such as personal transport, inspection and logistics.
In the automotive sector, the team will look at autonomous driving and advanced driver assist technologies such as personalised mobility pods. Meanwhile, the Inspection Flagship will concentrate on inspection and mapping from moving platforms, such as those used during the inspection of nuclear and chemical plants, roads and rail. This will create challenges, as the areas to be mapped will range from areas of a few metres to hundreds of kilometres, and the robotic systems will have to learn how to move in order to create maps of their surroundings. If successful, developments here could have a huge impact on the costs of decommissioning in the UK’s nuclear industry and the North Sea.
As the research progresses and the technologies within it develop, they will be rolled out virally to all of the projects across the program – for instance, a new surveying system could be rolled into mapping as well as the mobility pods in order to find out which parts of it are robust and which will need further development work.
‘As always with robotics, it is the unusual events that are intrinsically difficult to solve as they don’t happen very often,’ said Prof Newman. ‘Another challenge is for the the robot to have some sense of its own performance.’