A new call for partnerships, Safeguarded AI funding and a peek at our robotics programme
What’s new at ARIA…
We’ve launched a funding call for our second programme — Safeguarded AI. Backed by £59m, this programme aims to develop the safety standards we need for transformational AI. Find out more and apply here before 28 May.
Our latest programme thesis, ‘Precision Neurotechnologies for Human Therapeutics’ is out for feedback. Read it and engage with it here.
We are accepting full-length proposals for our ‘Scaling Compute’ programme. Find everything you need to apply here and submit your proposal before 7 May.
We’re looking for Activation Partners to design and deliver new, bespoke activities focused on science entrepreneurship and translation within our opportunity spaces. Learn more here.
In early June we’ll be launching our next programme ‘Robot Dexterity’, which aims to radically enhance robotics capabilities. Hear more from Programme Director Jenny Read in the Q+A below.
From stereoscopic vision to robotic dexterity – Q&A with Jenny Read
Your opportunity space is focused on robotics, but before joining ARIA, you were in vision science. What prompted the shift and why robotics?
While vision science is technically a different field, I've always been interested in robotics.
Before ARIA, I spent 20 years studying stereoscopic vision, which is traditionally thought of as a very complex and demanding computational ability, reserved in nature only for humans and other highly intelligent animals. When I learnt that even an insect as simple as the praying mantis has stereoscopic vision, I began to think about ways to produce a less complex and more cost effective form of stereoscopic vision in machines, and what sort of robotic applications would be appropriate.
The concept of insect-inspired robotics was part of my original pitch when I applied to be a Programme Director. But after I was offered the job, ARIA encouraged me to think about robotics more broadly, in pursuit of defining an opportunity space that could yield maximum impact.
How did you go about immersing yourself in a new field? What was the biggest challenge?
It’s been brilliant – I’ve absolutely thrown myself into it. I started out by meeting and speaking to as many people as I could. I’ve met over 150 people so far right across the robotics community in the UK and beyond, through university visits, workshops, one-to-one calls and conferences.
I've really enjoyed having that outsider perspective. It’s also reinforced that robotics is an incredibly wide field. Even if I were a subject matter expert in one area, I would still be an outsider in others. That’s provided a really helpful perspective when it comes to shaping a multidisciplinary programme, and bringing together scientists from different sectors.
What's been the most surprising moment of the discovery process so far?
I was surprised by how little hardware (e.g arms and end-effectors) have advanced in the past 50 years relative to the advancements in compute and software. I was also surprised by just how hard it is to innovate in robotics, and how the lack of common standards and an integrated software stack really complicates things.
What is the technical breakthrough you’re targeting through ‘Robot Dexterity’? Why is it important?
Dexterity in robotics remains an unsolved problem. Today’s robots have very poor information about their environments and about their own bodies. They have great visual abilities, but they don’t have anything like the other sensory abilities that biological organisms are covered in, which inform how we move through the world.
Imagine trying to tie your shoelaces, for example, but not having sensation in your hands, maybe because you've got pins and needles, or because your hands are really cold. When sensory footage is disrupted, everyday tasks are really hard.
There is growing interest in exploring improved sensing for robots, given the enormous untapped potential of robotics. Different technologies are being explored, but there’s no clear answer yet on how to achieve general dexterity. I'd like to fund research from different organisations exploring these technologies to see if we can identify the solution.
You’ll shortly be launching your programme solicitation. Who are the kinds of people you want to hear from and apply for funding?
In early June, I'll be launching a call for short concept papers, before asking people to set out their vision in a full length application. In terms of who I’m looking to hear from, it could be roboticists working in startups, public labs, big corporates, universities, or even individuals. Ultimately, I’ll be looking to build a portfolio of projects that forms a unified programme aimed at solving one or possibly two or three specific tasks in robotic dexterity.