We're trying to work out why it is hard to use autonomous technology in the maritime, air and land domains, and then doing projects that are impactful to try and alleviate some of those hurdles.
The collaboration with the York team came about through TAS’s previous CEO, Professor Jason Scholz, who encouraged the collaboration that Rachel then set up. AAIP is one of the primary international collaborators for TAS’s Assurance of Autonomy activity.
The AAIP team at York are doing world leading research and are at the forefront of the field that they're working in, and that we're also seeking to work in. We don't have any projects specifically looking at how to actually do assurance. For example, what should those assurance methodologies look like? And certainly not to the level of detail that the
AAIP have produced. For us, it's a no brainer to collaborate with such an organisation, especially one that's also funded for the public good.
The AAIP team delivered a series of three public webinars talking through the AMLAS guidance, and the ethics of assuring AI. Rachel also visited the AAIP at the Institute for Safe Autonomy in 2022.
The idea was to draw on the expertise of the AAIP team, and especially John McDermid, and find topics that were of interest to our TAS stakeholders. It was simultaneously more exposure for the AAIP team but also a really good way of giving our stakeholders access to world leading expertise.
Rachel emphasises that it would be good to collaborate further. She identified how the AAIP’s technical guidance does not duplicate the more general guidance produced by TAS.
Much of TAS’s work is about the application of technology (‘test and evaluate’ programmes) and the accelerated use of academic outputs; a trajectory which may offer clues for the next phase of the AAIP. Another area of potential collaboration is upskilling, specifically short courses to equip operators with the capability and understanding they need to use the technology in a safe and trusted way.
Centre for Assuring Autonomy