METRO Now Testing Out Self-Driving Buses for Houston

METRO NOW TESTING OUT SELF-DRIVING BUSES FOR HOUSTON Preliminary tests of the new autonomous buses METRO is studying for potential use in Houston have gone well so far, agency board member Sanhjay Ramabhadran said at a luncheon in Cy-Fair on Tuesday — “although,” he added, “they tend to overheat in the Texas weather.” The next test phase, METRO spokesperson Jerome Gray tells abc13, “will be on the campus of Texas Southern University with a small bus at slow speeds.” If it goes well, a prototype could then graduate to real traffic. Last January, the U.S. Department of Transportation named Houston one of 10 official proving grounds for driverless buses “in part for its miles of HOV lanes that could easily work as testing lanes.” [abc13] Photo: METRO

20 Comment

  • Great, driverless buses combined with post-millennials stepping out into traffic while intently watching videos on their phones. What could go wrong?

  • I like my bus drivers, to be honest. Is there a real tangible benefit to replacing bus drivers? One accident and lawsuit would probably offset 10yrs of salary savings on this technology.

  • Yet another reason to stay in the tunnels.

  • Random: What can go wrong? A lot less than the same scenario with a human driver. Self driving cars are coming like it or not.

    I took my Tesla to arizona from San Diego and took the wheel or pedals a total of 5 minutes maybe. That would have been 0 minutes of regulators would catch up with tech.

  • @ Turning_Basin: Human-driven buses also incur at-fault accidents, and the evidence thus far with self-driven vehicles seems to indicate a better track-record than human-driven vehicles. Now of course, METRO carries insurance for that sort of thing in order to limit the upper bounds of their exposure to liability, with or without drivers being in the buses. However, if there is a driverless bus and the technology is at fault and is licensed from a third party which indemnifies METRO in the event of that problem, then that might actually protect METRO even further and/or bring down the cost of their insurance.
    I for one look forward to this tech being deployed. If cost-savings allow METRO to put smaller buses on lower-ridership routes and to keep to schedule better, that’d be excellent for them and for the public. They would probably need an additional security presence, however.

  • This technology will rearrange Metro and their ideas. Driverless BRTs might be the poor-agencies version of a massive-area light rail system.

  • niche: could you post said evidence? that autonomous buses incur less accidents than human controlled buses. I am under the impression at best preliminary tests at best have been underway and that your statement is anticipatory not based on fact. then there is vegas where its one 20mph autonomous bus couldn’t make it 2hrs avoiding a situation a human driver would have reacted to

    also this article below makes a few nice points.

  • I am sure there is the tech out there, but how would a bus know a person is waiting at a stop to get on the bus (for a regular street route). I mean you don’t want the bus to stop at every stop or even when it “senses” a person at the bus stop (multiple routes using same stop, pedestrian walking past, etc). That’s the interesting wrinkle to see how they solve that with a self driving bus on a regular route. Now self driving Park and Ride route type busses, that’s “easy”

  • Good point, Robert Rutherford.
    Back when there was a bus stop on San Jacinto beside Fiesta, I was among several people at the bus stop when a Metro bus just breezed past us. One of the “regulars” advised me that the bus stopped only if someone was standing by the sign. The drivers knew that the people who were hanging out in the bus shelter had no intention of catching a bus. I doubt that a driverless bus will be able to make that distinction.
    Also agree with The Niche that a bus driver adds a welcome measure of security. One of the downsides of our rail system is the shenanigans that sometimes erupt on rail cars and the lack of an authority figure to deal with them.

  • @Robert Rutherford: Any time you want to ask the question: “How would a driverless car (or any other robot) do something people do?”, ask how people do it. The driverless car will do something very similar. Bus driver brains are trained to recognize the shapes of people. When they recognize those shapes in certain poses at a bus stop they stop the bus. Same thing for a robot.

  • To all those holders-on-of-times-gone-by who fear the future in which vehicles will be controlled by machine rather than by human, I offer this little tidbit: our roadways have been at the mercy of automation for quite some time. Have you ever driven on a street with an automated traffic cop? Traffic cops have been replaced in favor of traffic lights a long, long time ago.

  • @ Turning_Basin: I did not say that self-driving buses have (present tense) a better record than human-driven buses. The track record with self-driven buses is indeed too limited to make that specific claim, however I believe it reasonable to infer that if self-driven passenger vehicles (e.g. Chrysler Pacificas) can achieve that milestone, that buses probably can too. What you’ve done there is you’ve committed a straw-man fallacy. Please don’t do that anymore.
    Furthermore, the CNN article you linked to regarding the incident in Las Vegas seems to implicate the human-driven delivery truck as being at fault for having backed-up into the bus, which, having sensed danger, had already come to a stop. The city government, police, and witnesses all acted or spoke in ways that corroborated the same story and there was no indication that that story was challenged. Based on that, it is not clear that you read the article that you linked to. Please don’t do that anymore.
    The Nashville Scene article, from its “Pith in the Wind” section, is comprised of five paragraphs, but the third and fifth of those are two words and three words respectively. The author’s intent is to defend capital expenditures on light rail. They cite a five-paragraph excerpt from a different article that was discussing semi-autonomous vehicles, which are not germane to the subject at hand. Please don’t cite such sources anymore.
    @ Robert Rudford: There are many ways that the interface might work for the pickup/drop-off part of bus routing. It could be app-based (like Uber or Lyft) or each bus stop could have basically just a button connected to a microcontroller with an antenna on it. If a particular stop is host to multiple routes then there can be multiple buttons, with POS functionality to make sure that mischievous riders couldn’t just hit all the buttons without also first paying for all the rides. This pretty low-level PLC programming, by the way. An Internet-of-Things (IoT) enthusiast could probably figure it out pretty easily just on their own.

  • @niche, “the evidence thus far with self-driven vehicles seems to indicate a better track-record than human-driven vehicles”. your “belief to infer” is laughable. give me your evidence or take your wishful pseudo-intellectualism elsewhere.

  • “having sensed danger” i.e. its algorithm was only able to sense an object and not a situation which is to my point, that a driver would have anticipated the situation and continued to respond (honk, reverse, etc.) . And were you able to assemble a committee to determine how to address “citation” on a blog? (which the i included the article link and did not “cite” the article) when the prompt to its inclusion is that it makes a few nice points. your inability to think past your inevitable punch line of “don’t do that anymore” is really telling of your character.

  • The technology has come a long way, but it is a pretty big leap to go from having drivers to no drivers at all. The current technology seems like it is well suited to taking over the more routine, uneventful driving tasks with a human driver taking over for more complicated, safety-sensitive, unexpected situations. There are lots of tasks that computers can be better at than humans, but it is undeniable that humans are still better at improvising and making sense of unexpected situations.
    I think it makes more sense to have computer-assisted driving for a while before making the jump to having no driver at all. This has the added benefit of giving the computer more learning samples as it observes real-life driver behavior in complicated situations, and it buys time for wider adoption of technology in cars which helps because computers have a much easier time driving alongside other computers than driving alongside humans.

  • Have any readers here taken the northbound exit off of 59 to the West Loop lately?

    If some hypothetical autonomous driving system could navigate that tangle of past and present lane markings, then they might just have something to build on.

  • Argh. Southbound 59 exiting to 610 North…

  • The way that these things will identify legit passengers will probably use facial recogniton and/or being able to hail a bus once at a specific stop, or even anywhere on a route, using a phone app.

    The whole thing will become part of a creepy Matrix scene that everyone will be comfortable with.

  • More people out of work. How grand. Thanks Metro, you can pick up your homeless former drivers with your driverless bus. How refreshing

  • I welcome driverless technology. Metro will find a way to screw it up though.