Edinburgh Napier University

  In controlled settings such as factories, robots are able to achieve many tasks efficiently and accurately. However, it is still a challenge to design robots that operate in unstructured, dynamic and outdoor environments without human supervision (i.e. are not remote-controlled). In such settings, \textit{unpredictable changes} can occur that can render the \textit{robot ineffective at best} and in the worst case, useless.
Robots must therefore be able to \textit{autonomously adapt} previously learned behaviours to new tasks and settings if they are to be exploited in more challenging environments. Using a quadrupedal mobile robot the size of a small dog, the goal of this project is to identify the weaknesses in current state-of-the-art learning methods in this setting, using a mixture of simulation and physical testing, and then evaluate potential solutions. This will lay the foundations for a future ambitious proposal to EPRSC in conjunction with a team of ecologists that proposes to utilise robots in the wild for long-term monitoring and conservation activity in Scotland.