#1 cleantech news, reviews, & analysis site in the world. Subscribe today. The future is now.

Autonomous Vehicles

Published on November 5th, 2015 | by Michael Barnard


Tesla Has The Right Approach To Self-Driving Cars

November 5th, 2015 by  

Tesla recently released its Autopilot mode for its cars. It has a fundamentally different intellectual approach to autonomy than Google’s, and it’s superior.

One of my backgrounds is robotics. I spent a year digging my way through PhD theses from robotics programs around the world as I worked on a startup idea for specific applications of swarm-based robots. We got as far as software architecture, simple simulations, 3D modelling of physical robots, and specific applications which had fiscal value. I have some depth here without pretending to be a roboticist, and I’ve continued to pay attention to the field from the outside.

So I feel comfortable in saying that, in general, there are two approaches for robots getting from Point A to Point B.

  • The first is the world map paradigm, in which the robot or a connected system has a complete and detailed map of the world and a route is planned along that in advance accounting for obstacles. Basically, the robot has to think its way past or over every obstacle, which makes for a lot of programming.
  • images

    Yes, that’s a cat in a shark costume riding a Roomba.

    The second is the subsumption architecture paradigm, in which a robot is first made so that it can survive environments it will find itself in, then equipped with mechanisms to seek goals. The robot then, without any idea of the map of the world, navigates toward Point B. The robot is robust and can stumble its way through obstacles without any thinking at all. The original Roomba vacuum cleaner was a pure subsumption beast.

Obviously, both have strengths and limitations and obviously, at least to me, a combination is the best choice, but it’s worth assessing Tesla’s vs Google’s choices based on this.

LiDAR-2Google is starting from the full world map paradigm. For one of its cars to work, it needs an up-to-date centimetre-scale, 3D model of the entirety of the route it will take. Google’s cars are ridiculously non-robust — by design — and when confronted with something unusual will stop completely. Basically, all intelligence has to be provided by people in the lab writing better software.

Why would Google start with this enormous requirement? Well, in my opinion without having spoken to any of the principals in the decision, it’s likely because it fits their biases and blindspots. Google builds massive data sets and solves problems based on that data with intelligent algorithms. They don’t build real-world objects. And the split I highlighted above in world map vs subsumption paradigms is a very real dividing line in academics and research around robotics. It was very easy for Google and world view robotics researchers to find one another and confirm each others’ biases. Others assert that Google is taking a risk-averse approach by leaping straight to Level Four autonomy, and while I’m sure that’s a component of the decision-making process, I suspect it’s a bit of a rationalization for their biases. It’s also being proved wrong by the lack of Tesla crashes to date, but it is early days.

To be clear, Google cars can do things Teslas currently can’t, at least in the controlled prototype conditions that they are testing. They can drive from Point A to Point B in towns and regions that Google has mapped to centimetre scale, which is basically areas south of San Francisco plus a few demo areas. You can’t get in a Tesla, give it an address, and sit back. These are clear performance advantages of the Google model over current Tesla capabilities, and while not trivial, are enabled by the world map model.

Tesla, on the other hand, is starting with the subsumption model. First, the car is immensely capable of surviving on roads: great acceleration, great deceleration, great lateral turning speed and precision, great collision survivability. Then it’s made more capable of surviving. All the car needs to drive on the freeway is knowledge of the lines and the cars around it. Then it adds cameras to give it a hint about appropriate speed. It has only a handful of survivability goals: don’t hit cars in front of you, don’t let other cars hit you, stay in your lane, change lanes when requested, and it’s safe. Because of its great maneuverability — survivability — it can have suboptimal software because it is more able to get out of the way of bad situations. And it has human backup.

And if that’s where Tesla was stopping, everyone who is pooh-poohing its autonomy would be basically correct. But Tesla isn’t stopping there.

Tesla is leveraging intelligent real-world research assistants to put focused, experienced instincts into its cars. They are called the drivers of the Teslas. Every action the Autopilot makes and every intervention a driver makes is uploaded to the Tesla Cloud, where it’s combined with all of the other decisions cars and drivers are making. And every driver passing along a piece of road is automatically granted the knowledge of what the cars and drivers before them have done. In real time.

So, for example, within a couple of days of downloading, Teslas were already automatically slowing for corners that they took at speed before. And not trying to take confusingly marked offramps. And not exceeding the speed limits in places where the signs are obscured.

Within a couple of days of being available, the first people Cannonballed across the USA in under 59 hours with 96% or so of the driving done by the car. Given Google’s requirements, they would have had to send at least two cars out, one or more with a hyper-accurate mapping functionality, then a day or a week later, when the data was integrated, the actual autonomous car. And there would have been no chance of side trips or detours for the Google car. It literally couldn’t drive on a route that wasn’t pre-mapped at centimetre scale. But the Tesla drivers could just go for it.

People are driving Teslas on back roads and city streets with Autopilot, definitely not the optimum location-only situations that others claim Tesla is limited to. And Teslas haven’t hit anything; in fact, have been recorded as avoiding accidents that the driver was unaware of. Survivability remains very high.

Tesla cars are driving themselves autonomously in a whole bunch of places where Google cars can’t and won’t be able to for years or possibly decades. That’s because Teslas don’t depend on perfect centimetre scale maps that are up-to-date in order to do anything. Subsumption wins over world maps in an enormous number of real-world situations.

Finally, Teslas have a world map. It’s called Google Maps. And Tesla is doing more accurate mapping with its sensors for more accurate driving maps. But Teslas don’t require centimetre-scale accuracy in their world map to get around. They are just fine with much coarser-grained maps which are much easier to build, store, manipulate, and layer with intelligence as needed. These simpler maps combined with subsumption will enable Teslas to drive from Point A to Point B easily. They can already drive to the parkade and return by the themselves in controlled environments; the rest is just liability and regulations.

The rapid leaps in capability of the Autopilot in just a few days after release should be giving Google serious pause. By the time its software geniuses get the Google car ready for prime time on a large subset of roads, Teslas will be able to literally drive circles around them.

Tags: , , , , , ,

About the Author

Mike works with startups, existing businesses and investors to identify opportunities for significant bottom line growth in the transforming low-carbon economy. He regularly publishes analyses of low-carbon technology and policy in sites including Newsweek, Slate, Forbes, Huffington Post, Quartz, CleanTechnica and RenewEconomy, with some of his work included in textbooks. Third-party articles on his analyses and interviews have been published in dozens of news sites globally and have reached #1 on Reddit Science. Much of his work originates on Quora.com, where Mike has been a Top Writer annually since 2012. He's available for consultation, speaking engagements and Board positions.

  • It almost seems as if the author only researched Tesla and wanted to write about it, without even checking how the Google Self-Driving car really works. His whole article is based on the fact that Google requires a map inch by inch, which isn’t true. It does use navigation at the level TomTom does too, but doesn’t require it to be inch by inch. The power of the Google car is that it sees all objects around it, not only the other cars on the road. It predicts when pedestrians will cross the street, a child will run after a ball hitting the street and I’ve even seen a video of a car that didn’t start driving over a railroad crossing until it saw that there was a free stopping space for the traffic light across the road. That is where you need all that intelligence in the software for. I’m actually afraid that with Tesla’s current model, an accident could happen far more easier and would put all self-driving and autonomous driving projects in a bad light.

    • GCO

      You just have been proven right.

  • James Adams

    I do not want a self driving car. Too much liability and complexity. Just another useless gimmick to require maintenance and repairs.
    Remember: Just because you can, does not mean you should.
    If I can’t drive, I will take a bus, train or plane.

    • Tommy Czerwinski

      Just because you don’t want to doesn’t mean you have to, they aren’t going to make cars where it is always in autopilot mode and you can’t control it. They will most likely give you the option. With that said I think that driverless cars will revolutionize this world and will benefit everyone while being in their car. You would be able to get work done, eat, text, and even go out for a drink or two and have your car drive you home safely and preventing a possible DUI. I wouldn’t completely rule it out if I were you because this will be a relevant decision for you very soon.

      • James Adams

        Good points. But what we are seeing with other “safety” features is that they will be standard. Which means there is another system that you will have to pay for; even if you elect not to use it, you still bought it. People are already abusing the collision avoidance systems now. Saying “I can text/eat/read and drive and not worry about hitting someone”.

      • James Adams

        True, but I also don’t want to pay for a feature that I will not use.
        Nor do I want a driver that is not paying attention to the road, thinking that the autopilot is in charge and hitting a pedestrian or cyclist. The woman that was driving a Tesla and rear ended a truck on the highway is one bad example of autopilot gone wrong. I think that Tesla is placing the blame on the driver….

    • Jerry3130

      The whole point of autonomous car is that you don’t need to own it, and let it sit in the garage 95% of the time anymore. It will be right there when you need it, at a fraction of the cost of a cab.

  • Zdenka Micka

    Tesla and Google should merge. Best of both worlds

    • Hell no. Please keep the Google noses out of my life. I will never buy such a ‘Googla’, as I don’t trust them one bit to respect my privacy.

    • Frank

      Nah, better if they compete, hopefull stealing each others best ideas along the way.

  • Kirk Hilles

    I respect Tesla’s approach because they put it out there. People are using it now on live cars. Oh sure, Google has logged in a million miles… going around in circles around a 12 sq mile city. They probably have 30 employees updating data continuously on that city, noting every time a traffic light or stop sign moves slightly and probably uses 10 TB of data for that small city.

    I really don’t like Google’s all-or-nothing approach. I mean, you really believe you can get it 100%? Really? Handle every poorly marked road? Every crazy traffic situation, every parking lot, one way street, every weather condition (night, rain and snow), etc? No. I understand Google’s argument about the “hand off”, but they are wrong. Its VERY feasible to have a car drive Interstates only. No traffic lights, no cops directing traffic, just monotonous, maintain distance with the car in front of you driving. THAT should be their goal. Get that 100% and then you can easily give people a 5-10 minute warning about when the vehicle will be leaving the interstate thereby eliminating that issue (or just have it automatically pull over).

    Do that and continually improve it for 20 years and THEN we can talk about fully self driving. Google has been known to kill projects… I’m not convinced they just won’t “throw in the towel” 10 years from now because they can only get it a 95% level and need to hit 100%.

    • JH

      Goog i s doing nicely with their solution and have the figures to prove it.

    • freethinker

      it’s early game still; wait until the teslas start having accidents. it will happen. it is unavoidable with learning software. perhaps this is what google is afraid of with their internal development.

    • Tommy Czerwinski

      I like both approaches and how different they are. It kind of tackles two different monsters at the same time. I think if these two companies combined their technologies it would help further the process of making driverless cars a reality. I do agree with you though on Tesla’s approach being the better of the two, and like how they are taking things one step at a time and google is doing it all at once. I think they should just perfect each feature as it comes along then move on to the next. Google should follow tesla’s model but using their own technology cause it sounds like they are a lot further along from a technical stand point but it doesn’t pan out that well as tesla’s does on the actual roads.

  • Mikgigs

    it seems that everybody wants to give hide “smart” opinion on the topic. this site becomes a tesla forum. How the cat on roomba is related to?

    • Peter Waegemans

      Agreed. This site needs to chill on the Tesla adoration.

      • I don’t think the point is Tesla adoration. The point is evaluating to approaches to the cutting-edge of personal transport. Mike has extensive background in the field and can actually offer a thoughtful evaluation based on deep knowledge. I don’t think he would care if it were Tesla or Nissan pioneering autopilot, and I certainly wouldn’t. Should we not cover the leading-edge of electric cars and autopilot simply because Tesla is at the front of the pack? That just doesn’t make any sense to me.

        I understand if some people have an issue with Tesla/Elon and don’t want to read about them, but then why open an article about them?…

        • Peter Waegemans

          “Should we not cover the leading-edge of electric cars and autopilot simply because Tesla is at the front of the pack?”

          Yes absolutely. What Tesla is doing is amazing. Though I’m not a car enthousiast, I welcome the fact that they’re clearly speeding up the change towards EV’s for the masses.

          My point is that many Tesla related articles on this site are clearly about the car and less about clean technology, which is what this site is all about.
          Now, I have come to understand you also run a site called EV Obsession, which is dedicated to EV’s. So, if I may boldly suggest some moderation on what belongs on CleanTechnica and what belongs on EV Obsession, I’m pretty sure the nagging will stop.

          • Ah, that argument. Not sure how much you know of my background, but for a short summary:

            I’ve been car-free for 11+ years and used to work promoting bicycling, transit, and walking (got a master’s degree in that as well). I don’t think anyone can claim that we are going to cut our addiction to automobiles in the next several decades… as much as that might be good for us.

            Anyone who acknowledges this and looks at the global warming and pollution challenges in detail can see that we need to electrify transport. We started covering EVs hesitantly, and I wasn’t into them, but as I came to realize their importance, it became a central focus of our site.

            The underlying need for EVs to gasmobiles is the reason we cover them so much. And I guess you can see that Tesla is obvious the leader in this realm.

            We do put stuff on EV Obsession that we don’t put here, just like we put stuff on Solar Love (which I also founded) that we don’t put here. But the stuff that seems to fit our vision for this site still goes here as well.

            Hope that clears things up.

  • serge delinois

    I think the difference in approach is also based on different goals. I do believe that google’s goal is to become the autonomous uber. They already have the pod cars and all they need is to complete the software. Just imagine 10,000 little pod cars driving through New York City ferrying people back and forth. Mapping the city is easily done by Google and the pods would never have to leave the mapped area.

    On the other hand Tesla sells cars to people and people want freedom to go anywhere. So their software is geared towards maximizing the benefit to the driver anywhere they go. Even if it’s only good 90% of the time.

    • Nice points.

    • Short term vs long term. Tesla will achieve full autonomy but is doing it incrementally. There will be Uber Teslas without drivers. And I think it will happen before Google actually gets something into production despite starting later.

      • serge delinois

        Having driven my Tesla’s with autopilot I can tell you it’s nowhere close to what I’ve seen from google. Google already has self driving cars driving around logging millions of miles completely autonomously.

        I’m a Tesla owner and a huge fan but you need to think logically and not emotionally.

        • I think you’ve sort of missed Mike’s argument. Maybe not, but I’ll try to repeat it succinctly in case:

          Google’s approach requires knowing every street the car will drive on to a very precise degree, and know every change as soon as it happens if the cars are actually going to be useful. It may be somewhat flexible, but not flexible enough to get somewhere if its maps are messed up. Whereas Tesla’s approach is to build cars that have decent enough (but not nearly as precise) maps to have an idea of what to do, good enough “eyes” and software to respond to unforeseen challenges, and the ability to learn from user feedback on the go.

          Imaging a time when Google has every street mapped out so precisely and updating so frequently that a car can take you wherever you want to go… well, that’s much harder to do than imaging a point where Teslas have good enough maps, sensors, radars, software, and user feedback to get you from one place to any address you put in as the destination.

          Then there’s also the issue of LIDAR being very expensive, but I think that’s another matter….

          • serge delinois

            I agree but as per my other post I don’t think they have the same goal in mind. Tesla wants to automate cars that people drive and I believe Google want to create automated Uber and I think they are very close to doing it.

            Let’s say they map out San Francisco a pretty small city in terms of square miles. Then they deploy google uber with 1000 of their pod cars. They could take over the whole taxi/ride sharing business. Those cars never leave the city except for driving people to the airports and stuff.

            That’s a business that can be done in < 5 years. Run that business for a year to iron out the small details and push it out to every major city as fast as you can build the cars.

          • I think Tesla has the same basic goal as Google in the long term, just a route that goes through consumer cars. But yeah, that is certainly up for debate.

            And I’m not an expert on the subject, so I’m not going to say “this company” will have an Uber-like fleet in big cities first, but Mike’s argument makes a ton of sense to me, and the little bit I know about this makes me think Tesla will. But yeah, don’t take my opinion as a very noteworthy one! Seriously. I’m just enjoying the show for now and trying to learn a bit about it all.

          • Brooks Bridges

            I can actually see a blend of the two approaches: Google part is in control until something it can’t handle pops up and then the Tesla approach takes over momentarily, then back to Google.

          • Jerry3130

            “Imaging a time when Google has every street mapped out so precisely and updating so frequently that a car can take you wherever you want to go….”
            I think Google already have precise maps on almost all the street in America, and they only need to download the map that’s 100 mile within where they are. It take more than 1 hour for the car to drive outside that range.

      • freethinker

        maybe Google buys Tesla?

    • Peter Waegemans

      That’s exactly how I see (parts of) the world in 10 years. You pay a certain fee each month and you Have access to a (probably shared) Google cab in every major city.

      Tesla ao will probably focus on people who prefer having their own car. Something that might become extremely expensive in a world where driving becomes obsolete.

    • Johnny Le

      Their goal is the same; their approach is different. Tesla will sell half of its model 3 to people and half to be like uber. At this point the car can open the door by itself, know where the superchargers are, and can charge by itself. All it needs is to drive to a supercharger on its own. I’m very sure Tesla will get there before Google.

      The nice thing about Tesla’s approach is that it becomes useful right from the start. It’s like a kid asking Tesla and Google to help with his math homework. Tesla says “sure” and helps while Google says “sure, but let me get my math degree first, then my teaching degree, and then I’ll help.” By the time google has everything in place, the kid already graduated from college.

  • JamesWimberley

    Interesting. You, plus the news from the Teslas on the road, have shifted me from a pro-Google stance to at least neutrality.

    Neither side seems interested in seeding the road system with millions of tiny IoT navigation beacons. This could be done quite cheaply, and supply a lot of useful information to the autopilots. You could have hazard beacons, dropped in an instant by first responders at accidents, etc.

  • Aditya

    I think the difference here is quite clear. Tesla is going for incremental change with their Autopilot, improving it every week if not every day. Google is aiming for the finished product directly, obviously more time consuming.

    With more Tesla vehicles joining the streets and mapping them for the company, mapping of roads and adding nuance to the driving is only going to speed up. But it will be interesting if and when they can make the jump from Semi Autonomous to Autonomous driving.

    The future is exciting!

    • freethinker

      to sum it up; tesla is crowd sourcing improvements to their world model with good enough functionality to enable those improvements to be possible.

  • Wilibald Oplatek

    Perhaps the author could disclose his information sources about the internal workings of the Google autonomous vehicles. As far as I know, Google has one of the largest traffic images databases so their neural networks can’t be that far behind Mobileye’s (Tesla’s). Claiming that Google relies solely on the world map sounds like an uninformed oversimplification to me.

    • Stephanie Campbell

      Google gives you a great Chance to make 18648$/2weeks at your home.If you are some intelligent you make many more.I am also earning many more, my relatives wondered to see how i settle my Life in few days thank GOD to you for this… You can also make cash i never tell alie you should check this I am sure you shocked to see this amazing offer… I’m Loving it!!! ☻ ▼ ▼ ▼
      ✔✔✔✔✔✔✔➤➤ http://www.dailycareeersplans.­&#116k

  • Aaron Lephart

    Whew, I was worried there was going to be a day without a TESLA related article. Bravo chaps!

    All the best,
    Aaron Lephart

  • Carl Raymond S

    I think you need to be clearer on what “self-driving” means. Sure, a Tesla can steer on a clearly marked road without maps, but would be destined to be stranded (or worse) if there was NO driver. The ultimate goal has to be autonomous driving, using BOTH strategies, and I’m sure Tesla knows this. Tesla must be laughing, knowing that every car they sell is a high-precision unpaid mapping-agent. By the time the cars are ready with the additional sensors and smarts, the maps will be too. There won’t be a road or road-lane in any country that Tesla sells to that hasn’t been traversed many times by the ‘fleet’.

    • Kyle Field

      Elon is clear at the distinction between autopilot and autonomous driving. They are very different – one requires a human, one does not.

Back to Top ↑