Jun 30 2009

Toyota Gets In The Game

Published by under Uncategorized
Comments: 9

In April of this year I wrote about Honda’s press release announcing that they had made some breakthroughs in the area of mental control of robots. On close inspection it seemed that Honda had not accomplished anything that other researchers weren’t already doing, so they were essentially announcing that they were in the game. Well, not another Japanese car manufacturer is in the game also – Toyota.

They announced that they have developed a system that allows a person to control a wheelchair with their thoughts alone.  The system allows a person to make the chair go right, left, or forward with thoughts alone. Curiously, in order to stop the chair a traditional puffer control is still needed (where the operator controls the chair by puffing into a tube).

This is nothing new – devising a system that can distinguish among three brain states is fairly crude. Obviously they could not make it distinguish a fourth to command the chair to stop. But Toyota needs their distinguishing feature – something to put in the press release to make their technology seem like a breakthrough. So, they tell us, theirs is the fastest system yet. While other systems take several seconds to process the thought commands, their system accomplishes this task in 125 milliseconds (thousandths of a second). That is fast enough to give the feel of instantaneous control. In fact the response time of the brain itself, from thought to action, is about 100 milliseconds.

While this is an incremental advance at best, it does sound like a legitimate advance. But more interesting is that we now have two large corporations investing in research into mental control of machines. They add to the research groups at Duke University and in New York who have also made headlines with their research.

The University of Pittsburgh has a suite of labs that are working on the various aspects of robotic mental control. At the Motorlab their research if focusing on correlating patterns of cortical activation with motor control in three-dimensional space. At the NTE lab, they are researching the interface between neural tissue and “smart biomaterials,” as well as neural tissue engineering. So they are working on the interface between brain and machine.

At the Rehab Neural EngineeringLab they are working on decoding the sensory feedback given to the brain from limbs during movement. This research could close the loop – allowing users to feel their prosthetic limb, which would greatly enhance control.

Although not mentioned on their site, there is straight neuroscience research which is discovering the neural activity that correlates with the brain’s sense of ownership over a body part – the sense that a body part belongs to you. So it is very plausible that in the future not only would a brain-machine interface allow a user to control a robotic limb, but also to feel it, receive information from the limb regarding its position and the tension on its “muscles”, and also feel as if it is part of their body. This is Six Million Dollar Man territory.

This is all very exciting. As the Neuro Rehab website says:

There is a rapidly growing brain-computer-interface (BCI) community working to develop methods of extracting and processing neural signals for directly controlling devices such as cmputers or prosthetic systems.

Now we also have large corporations like Honda and Toyota in the game – which is a sign that a technology is poised to move out of the lab and into the market. These companies say they have not plans to market any particular technology, they are just doing technology research.  This is clear from the crude devices they are demonstrating. But it means that we are beyond the basic university level research into developing technological applications.

We must also remind ourselves that while this technology is steadily progressing, and has already made interesting advances, it is probably years away from the simplest applications and decades away from any mature applications, like fully functional neuro-prosthetic limbs.

That is the one downside to tracking new and exciting technology in its early phase – you have to wait years to decades before the promised applications emerge. It’s like seeing an awesome movie preview a decade before the release date.

Share

9 responses so far

9 Responses to “Toyota Gets In The Game”

  1. jonny_ehon 30 Jun 2009 at 10:36 am

    Have you considered that the braking mechanism is separate from the brain scanning tech in order to make it more reliable/safe? If you’re headed towards an empty elevator shaft, and panic, you don’t want the chair to accidentally speed up! Think happy thoughts, FAST!

  2. Steven Novellaon 30 Jun 2009 at 11:17 am

    jonny – yeah, but that also tells you something about the technology.

  3. Bronze Dogon 30 Jun 2009 at 6:11 pm

    Always good to have more players on the field, especially corporations who have a direct monetary interest in making viable products.

    So, when does Ghost in the Shell come to a RL near us, again?

  4. HHCon 30 Jun 2009 at 10:50 pm

    Dr. Novella, I would like to say thanks to you for your frequent discussions about wheelchair patients and their neurological functioning. My mom is 88 and has not walked since August, 2007. She can walk today 35 feet with a walker and some assistance. It has been a complicated recovery but I know that I was more confident in my approach to the ordering physician and staff to make this happen thanks to your blog.

  5. Saorsaon 01 Jul 2009 at 6:52 am

    As a quadriplegic, I can’t wait to become a fully ambulatory cyborg :)

    Quick question Dr. Novella: I’ve seen chips implanted into muscles that can stimulate them to contract. Why hasn’t anyone implanted many of these in multiple muscle groups, interfaced them with a computer, and programmed them to perform organized, complex muscle contractions, like standing or fine finger movements? Hand control, or simply standing for 2 minutes to transfer could improve so many lives.

  6. Phospholipidon 01 Jul 2009 at 9:30 am

    @ Saorsa:

    G.E. Loeb at University of Southern California is doing just that with BION chips if I’m not completely mistaken.

  7. Calli Arcaleon 01 Jul 2009 at 10:16 am

    Talking of stimulating muscles reminds me of something I read recently. Chimpanzees are vastly stronger than humans, despite their muscles not being all that much different, in size or design. The reason is that chimps trigger entire muscles when they want to move, but we only trigger portions of muscles. Thus, we aren’t able to get the full performance out of a muscle. But we get something perhaps far more important — fine motor control. Being able to trigger just part of a muscle means we have extremely delicate control. This is why we can craft such sophisticated tools and such exquisite art, and why we can write (written language being perhaps the most powerful tool we ever devised).

    I wonder if this is part of the challenge in building computer controllers for muscles — the computer is likely going to stimulate the entire muscle, but our normal movements require much finer control.

  8. Phospholipidon 02 Jul 2009 at 4:16 am

    I think that depends entirely on what you mean by fine control. On the one hand (no pun intended), dextrous movement of the hand *can* be precisely controlled for certain tasks, while other require less accuracy, like holding on to a handle. On the other hand, leg movements are usually considered less accurate, but being able to somewhat stand up wouldn’t cut it – falling over would be a problem.

    I should admit my bias. I work in neuroprosthetics, so I’m not really impressed by direct muscle stimulation. If you’re interested, I recommend this article, though it’s also a bit slanted towards neural rather than intramuscular implants:

    http://www.ncbi.nlm.nih.gov/pubmed/17873417?ordinalpos=3&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_DefaultReportPanel.Pubmed_RVDocSum

  9. faseidlon 02 Jul 2009 at 4:31 pm

    I come at this from the perspective a computer engineer–both hardware and software–and a someone with a long-time interest in minds, machine, disruptive technologies, and evolution. What I find absolutely fascinating about this story (and the Honda announcement at the end of March) is the overall innovation trajectory we are on.

    It’s easy to look at a new technology and start thinking about how it might change the way we live. But I think it goes deeper than that. In a very real sense, technology innovation is changing the way we *are*.

    I blogged more about this a couple days ago:

    Brain/Machine Interfaces (BMI): Something to Think About
    http://faseidl.com/public/item/235534

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.