Tuesday, May 28, 2013

New Robotics and new opportunities

Here are the slides of my talk at the BARA Academic Forum for Robotics meeting Robotics: from innovation to service, on Monday 20 May 2013:

The key messages from my talk were:
  • The new wave of robotics represents a kind of Cambrian explosion in robotics: an exciting but also bewildering exploration of new forms, functions and materials. This explosion of diversity means that the New Robotics is not one kind of robot. Thus any kind of prediction about which of these will successfully evolve to become mainstream is more or less impossible.
  • There are two common myths: first, the waiting-for-AI myth: the idea that robotics is waiting for some breakthrough innovation in Artificial Intelligence, without which robotics is stuck. And second, the need-full-autonomy myth: the idea that fully autonomous robots represent some ideal end-state of the development of robotics; this is not true - instead we need robots and human-robot interfaces that will transition smoothly between tele-operation and semi-autonomy. We call this dynamic autonomy.
  • There are significant opportunities for innovation right now - underpinned by a significant head-of-steam of fundamental technologies from university R&D. I offer some examples for discussion, including companion robots, wearable robots and tele-operated robots with immersive tele-presence, perhaps making use of remote tele-haptics (although I claim no special insights). 
  • We need new and agile approaches to innovation. New kinds of research-industry partnerships and flexible, responsive pathways to commercialisation. Especially campus start-ups and incubators, nurturing post-docs as next generation entrepreneurs; and innovative modes of funding. We also need responsible and sustainable innovation.Haptocs 

Here are links to further information, and video clips, on the projects and robots highlighted in the talk:

Slide 10: The Cooperative Human Robot Interaction Systems (CHRIS) project
Slide 11: MOBISERV - An Integrated Intelligent Home Environment for the Provision of Health, Nutrition and Well-Being Services to Older Adults
Slide 12: Hand exoskeleton for post stroke recovery

Slide 13: Tactile Sensing - tele-haptics

Slide 14: Surgical Haptics
Slide 15: Search and Rescue - Disaster Response
Slide 16: Towards energy sustainability

Sunday, May 26, 2013

What is the single biggest obstacle preventing robotics going mainstream?

The question Robotics by Invitation asked its panel in May 2013, was:

What is the single biggest obstacle preventing robotics from going mainstream? It has been said that we are on the edge of a ‘robotic tipping point’ … but where, exactly, is this edge? And what’s holding us back?

Here is my answer:

It depends on what you mean by mainstream. For a number of  major industry sectors robotics is already mainstream. In assembly-line automation, for instance; or undersea oil well maintenance and inspection. You could argue that robotics is well established as the technology of choice for planetary exploration. And in human culture too, robots are already decidedly mainstream. Make believe robots are everywhere, from toys and children’s cartoons, to TV ads and big budget Hollywood movies. Robots are so rooted in our cultural landscape that public attitudes are, I believe, informed – or rather misinformed – primarily by fictional rather than real-world robots.

So I think robotics is already mainstream. But I understand the sentiment behind the question. In robotics we have a shared sense of a technology that has yet to reach its true potential; of a dream unfulfilled.

The question asks what is the single biggest obstacle. In my view some of the biggest immediate obstacles are not technical but human. Let me explain with an example. We already have some very capable tele-operated robots for disaster response. They are rugged, reliable and some are well field-tested. Yet why it is that robots like these are not standard equipment with fire brigades? I see no technical reason that fire tenders shouldn’t have, as standard, a compartment with a tele-operated robot – charged and ready for use when it’s needed. There are, in my view, no real technical obstacles. The problem I think is that such robots need to become accepted by fire departments and the fire fighters themselves, with all that this entails for training, in-use experience and revised operational procedures.

In the longer term we need to ask what it would mean for robotics to go mainstream. Would it mean everyone having a personal robot, in the same we all now have personal computing devices? Or, when all cars are driverless perhaps? Or, when everyone whose lives would be improved with a robot assistant, could reasonably expect to be able to afford one? Some versions of mainstream are maybe not a good idea: I’m not sure I want to contemplate a world in there are as many personal mobile robots, as there are mobile phones now (~4.5 billion). Would this create robot smog, as Illah Nourbakhsh calls it in his brilliant new book Robot Futures?

Right now I don’t have a clear idea of what it would mean for robots to go mainstream, but one thing’s for sure: we should be thinking about what kind of sustainable, humanity benefitting and life enhancing mainstream robot futures we really want.