Tomfoolery – Autonomous Motorcycling

Share this Article

Are you aware that BMW’s Highly Automated Driving prototype technology is capable of drifting a 2 Series coupe? Check out the video below, specifically from 0:58 to 1:06 and witness an autonomous stunt that’s cameo worthy in the next Fast & Furious.

You’ve probably heard about the forthcoming smart cars or have maybe even seen the driverless Google car being tested. While cars featuring basic autonomous capabilities are already in dealership showrooms, consumer availability of completely autonomous cars hasn’t arrived, yet, but the advent of that technology is gaining momentum. According to a Morgan Stanley research report, “completely autonomous cars are likely to be available before the end of the decade.”

autonomous-vehicleWhen completely autonomous vehicles do make it to market – become accepted, commonplace and finally indispensable – what happens to our precious motorcycles? It’s doubtful BMW, Harley-Davidson, Honda, et al., will quietly invite the demise of the two-wheeler, but how will the development of safety and autonomous technologies be applied to motorcycling? Are we destined to become nothing more than WALL·E characters riding around on communicating recliners?

Can’t happen. Won’t happen. Motorcycle operation is a complicated and delicate balance of man and machine that technology is incapable of reproducing, you say? I, too, scoffed at the notion, thinking that when this advanced technology arrives I will have already retired my riding boots. Then I came across this guy…

Jingjang Yi is an Associate Professor at the Department of Mechanical and Aerospace Engineering at Rutgers. His compendium of work includes the following titles:

  • Autonomous Motorcycles for Agile Maneuvers, Parts I and II
  • Autonomous Control of Riderless Motorcycles
  • Vision-based motion planning for an autonomous motorcycle on ill-structured roads
  • Vision-based navigation for an autonomous motorcycle on ill-structured roads
  • Balance control and analysis of stationary riderless motorcycles
  • Velocity field-based maneuver regulation of autonomous motorcycles
  • Trajectory tracking and balance stabilization control of autonomous motorcycles
  • Dynamic Modeling of Riderless Motorcycles

Co-author on many of these studies is Dezhen Song, Associate Professor in the Department of Computer Science and Engineering at Texas A&M University. Song provided me with the video below, shot in 2006.

Clearly, the video demonstrates the technology required to operate an autonomous motorcycle existed nearly a decade ago. Initial research focused on utilizing autonomous motorcycling technology for use in search and rescue missions in addition to military applications. According to Song, the benefits of autonomous motorcycles is that they can access about 98% of terrain which makes an autonomous motorcycle a good candidate for this type of operation.

A Great Recession later and government funding has dried up like a desert flood wash, leaving Song and his autonomous motorcycling collaborators without compensation. When Song approached major motorcycle OEMs, he discovered companies such as BMW were uninterested in pursuing autonomous motorcycle technology outside of safety purposes such as vehicle-to-vehicle communication addressed in my last editorial, The Future Of Motorcycle Safety Awareness Month.

Lacking any commercial viability, Song has broadened his approach to developing algorithms for all autonomous vehicles, using his autonomous motorcycle research for developing human motor-skill testing. He’s also tackling issues related to the deployment of autonomous technology.

“Basically, we’re trying to lower the barrier for human psychological acceptance because it’s hard to believe autonomous technology is reliable,” says Song. “And also lower the barrier for non-technical issues such as insurance. If a self-driving car hits someone, who’s going to be responsible for that?”

robot_ethicsSong is only scratching the surface of moral/ethical algorithm hell, a topic Patrick Lin addressed in an article for Wired magazine where Noah Goodall, a research scientist at the Virginia Center for Transportation Innovation and Research, presents this scenario:

If an autonomous car is facing an imminent crash with one of two motorcyclists: one who is wearing a helmet, and one who is not, a proficient algorithm accounts for the much-higher statistical odds that the biker without a helmet will die if struck, and therefore, in the name of crash optimization, chooses to strike the helmeted motorcyclist.

Is this the correct moral/ethical decision or an injustice? By choosing to strike the helmeted motorcyclist, the autonomous algorithm is penalizing the helmeted motorcyclist for being responsible and safety conscious. Meanwhile, the unhelmeted motorcyclist is spared, even though that rider is guilty of irresponsibility and lacks common sense self-preservation instincts.

Thinking about the infinite scenarios surrounding autonomous technology makes my brain hurt. When a dog runs into the street, does your autonomous motorcycle choose to plow the dog or ride you into a brick wall? An autonomous car must choose between striking a car with four passengers but better protection vs. striking a motorcycle with one passenger but less protection. Which vehicle gets hit?

Ugh, please pass the Vicodin.

Cost remains a problem, but like other technologies, complete autonomy will first be introduced on luxury vehicles and then filter to less expensive vehicles as prices decrease. I trust the timeframe from now until the ubiquity of autonomous cars and motorcycles is reached will provide ample opportunity to solve these technological/moral/ethical dilemmas.

How autonomous technology, and all the promises and problems associated with it, will affect motorcycles and riders will slowly play out before our eyes. Who knows. Future motorcyclists may one day view twistgrips to be as novel to them as kickstarters and drum brakes are to us.

For now, I aim to enjoy full control over the motorcycles I ride, accepting both risk and reward. But maybe, some day, I’ll be able to ride to the local dive bar, get sh*tfaced, then take a leisurely, helmetless and ultimately safer ride home than my current sober, properly outfitted self can. That’s me looking on the bright side of the future.

Autonomous_Terminator_Cycle

Get Motorcycle.com in your Inbox
  • Steven Holmes

    hey where did you get a picture of my riding buddy?

  • john burns

    funny I just got spam from Motley Fool saying next big stock is the guy who’s cornered the market in autonomous control, and it’ll be here much sooner than we think.

    • http://motorcycle.com/ Tom Roderick

      Maybe there’s more to it than just being spam?

  • Oslo Norway

    I need to step up the R&D on that Way Back Machine project…

  • Ser Samsquamsh

    A lot of people aren’t “driving”. They are providing the minimum AI to steer their vehicle mostly in the right direction. Myself included. When I have to drive my truck to work I’m pretty much on autopilot anyway. That would be perfect if I could put my feet up and get an extra 20 minutes of sleep!

  • John A. Stockman

    This tail-wags-the-dog approach to “making things safer” is the usual practice of working from the back end forward, instead of addressing the BIG issue: driver training, at least in this country, is appallingly non-existent. And the numerous distractions that take attention and eyes off the road is amazing to me. “TV” screens on the dash, oh sorry, info-tainment/radio/heating/communication screens that require you to actually look at them to operate them. No wonder people are crashing into each other. The push for autonomous cars is inevitable because no one wants to address the real issue. Until the process to get a driver’s license in this country is upgraded and taken more seriously, the average driver will continue to have a nonchalant attitude about driving proficiently. Yeah, I’m an idiot, how dare I think that. I’m with Glynn Kerr, anyone (within reason of course) who wants a drivers license and is physically able should be required to take a training course to get their motorcycle license, and operate a motorcycle for two years before they can get the training and license to drive a 4-wheel vehicle.

    • http://motorcycle.com/ Tom Roderick

      You, me and Kerr must be the three idiot amigos because I’ve been expressing the same opinion about driver training and motorcycle operation for years. But, I’ve also accepted that neither will ever happen in this country.

      • John A. Stockman

        The 3 idiot-amigos! I love that. I’ve been proposing and jumping up & down about that stuff probably as long as you have Tom. I was sending snail-mails to my representatives and pertinent officials way before e-mails, and now I do with e-mail. I’m not one to give up, but I know deep down you’re right about it never happening in America. When visiting and talking with my European friends that are motorcyclists, I know how training is approached in other countries and how much effort it takes to get a driver/motorcycle license in places other than the USA. They all wonder if it will ever get beyond go-through-a-few-traffic-lights/corners-and-parallel-park-for-me and here’s your drivers license. Motorcycle instruction is getting there, but there are large groups of riders and potential riders that think they don’t need it at all. The state-DMVs should start deferring anyone who wants a motorcycle endorsement to the various training programs, and stop doing those ride-through-the-cones exercises to get your motorcycle license. Any motorcyclist that realizes the need for proper training and a good kit knows why motorcycling (and even the licensing part of it) continues to be marginalized here, but no one wants to talk about the real root of it. But I will gladly accept our 3-Amigos moniker! Thanks for the good laugh.