Soft robots, Part 1 — introduction
There may be no other subject on which I’m so potentially biased as robotics, given that:
- I don’t spend a lot of time on the area, but …
- … one of the better robotics engineers in the world (Kevin Albert) just happens to be in my family …
- … and thus he’s overwhelmingly my main source on the general subject of robots.
Still, I’m solely responsible for my own posts and opinions, while Kevin is busy running his startup (Pneubotics) and raising my grandson. And by the way — I’ve been watching the robotics industry slightly longer than Kevin has been alive. 😉
My overview messages about all this are:
- Historically, robots have been very limited in their scope of motion and action. Indeed, most successful robots to date have been immobile, metallic programmable machines, serving on classic assembly lines.
- Next-generation robots should and will be much more able to safely and effectively navigate through and work within general human-centric environments.
- This will affect a variety of application areas in ways that readers of this blog may care about.
Examples of the first point may be found in any number of automobile factory videos, such as:
A famous example of the second point is a 5-year-old video of Kevin’s work on prototype robot locomotion, namely:
Walking robots (such as Big Dog) and general soft robots (such as those from Pneubotics) rely on real-time adaptation to physical feedback. Robots have long enjoyed machine vision,* but their touch capabilities have been very limited. Current research/development proposes to solve that problem, hence allowing robots that can navigate uneven real-world surfaces, grip and lift objects of unpredictable weight or position, and minimize consequences when unwanted collisions do occur. (See for example in the video where Big Dog is kicked sideways across a nasty patch of ice.)
*Little-remembered fact — Symantec spun out ~30 years ago from a vision company called Machine Intelligence, back when “artificial intelligence” was viewed as a meaningful product category. Symantec’s first product — which explains the company name — was in natural language query.
Pneubotics and others take this further, by making robots out of soft, light, flexible materials. Benefits will/could include:
- Safety (obviously).
- Cost-effectiveness (better weight/strength ratios –> less power needed –> less lugging of batteries or whatever –> much more capability for actual work).
- Operation in varied environments (underwater, outer space, etc.).
- Better locomotion even on dry land (because of weight and safety).
Above all, soft robots will have more effective senses of touch, as they literally bend and conform to contact with real-world surfaces and objects.
Now let’s turn to some of the implications of soft and mobile robotic technology.
Related links
- This series partially fulfils an IOU left in my recent post on IT innovation.
- Business Week is one of several publications that have written about soft robots.
- Kevin shared links to three more videos on robot locomotion.
- What I wrote about analyst bias back in 2006 still applies.
Comments
2 Responses to “Soft robots, Part 1 — introduction”
Leave a Reply
[…] will soft, mobile robots be able to do that previous generations cannot? A lot. But I’m particularly intrigued by two […]
[…] Earlier this year, I posted about robotics. […]