Mobile Autonomy Programme Grant: Safety, Trust and Integrity
This project is one theme of the EPSRC-funded programme grant Mobile Autonomy: Enabling a Pervasive Technology of the Future (grant reference EP/M019918/1) led by Paul Newman (Engineering Science), with co-investigators Ingmar Posner (Engineering Science), Niki Trigoni and Marta Kwiatkowska (Computer Science).
Vision: To create, run and exploit the world's leading research programme in mobile autonomy, addressing fundamental technical issues, which impede large-scale commercial and societal adaptation of mobile robotics.
Safety, Trust and Integrity Theme: As autonomous systems become widely adopted by the general public, potentially increasing the risk to society, we must ensure that the mobile robots we design are safe (e.g. the software does not crash while performing a parking manoeuvre) and their decisions trustworthy (e.g. the rescue-robot is able to precisely identify the location of a victim), without impacting time-to-market deadlines. Since mobile robots are controlled by embedded software, computer-aided tools to support their design must be used to achieve the required level of software quality assurance and device reliability. We will develop a model-based verification and validation (V&V) framework that has the power to guarantee that autonomous systems we build and deploy, meet the stated design objectives. These will range from safety and trust requirements, to energy efficiency, and will be integrated within the sensing, vision and planning technologies. The novel angle of this Theme is our goal to focus on quantitative models and objectives, and to ultimately generate correct-by-construction embedded software components from quantitative specifications.
3D−PhysNet: Learning the Intuitive Physics of Non−Rigid Object Deformations
Zhihua Wang‚ Stefano Rosa‚ Bo Yang‚ Sen Wang‚ Niki Trigoni and Andrew Markham
In 27th International Joint Conference on Artificial Intelligence and the 23rd European Conference on Artificial Intelligence IJCAI−ECAI. 2018.
VINet: Visual Inertial Odometry as a Sequence to Sequence Learning Problem
R. Clark‚ S. Wang‚ H. Wen‚ A. Markham and N. Trigoni
In AAAI Conference on Artificial Intelligence (AAAI). 2017.
DeepVO: Towards End−to−End Visual Odometry with Deep Recurrent Convolutional Neural Networks
S. Wang‚ R. Clark‚ H. Wen and N. Trigoni
In International Conference on Robotics and Automation. 2017.