Quantcast
≡ Menu

Exponential Technology and the Self-Driving Car‏

Mercedes Self-Driving Car‏

Photo Credit Ernest Dickmanns

On a crisp autumn day, two sleek Mercedes sedans smoothly joined the flow of traffic on French Autoroute 1 near Paris’ Charles de Gaulle airport. For hours the two vehicles drove along the highway in heavy traffic, passing other cars when necessary. Although no obvious signs distinguished them as unique, close observation would reveal that both were autonomous vehicles, driving themselves with human-level competency.

Self-driving vehicle technology is now beginning to establish itself in the market. Major automakers like Volkswagen have been quietly shipping driverless technologies in new production vehicles. Tesla Motors nears completion on a promised software update that will enable autopilot mode. But the two Mercedes in question – VAMP and VITA-2 – are not current production models, nor are they current research vehicles. Today they sit in a museum in Bonn, Germany, because the above drive happened in 1994 – over 20 years ago. We will use the surprisingly old story of the development of the self-driving car to illustrate an often-overlooked aspect of the development of enormously successful technologies – they can take a long time to become an “overnight success.” Such exponential technologies reach a point where they suddenly burst forth into the public eye. We call this point the “knee of the curve,” where the exponential goes from doubling small numbers to produce more small numbers, to doubling more substantial numbers that quickly become huge. Public attention understandably focuses on the breakout and scaling of advanced technologies. But reaching the breakout point requires building the necessary base, which can take a long time – much longer than people often realize.

Scientists had worked on the problem of self-driving cars for decades. Research pioneers produced crude driverless cars during the 1950s – but these early systems required help from modified roads. General Motors and RCA demonstrated an early system that used embedded cables in the roadways, tracked by magnets. This technology was never deployed due to the expense and difficulty of redesigning roads.

A new approach was needed. Enter Ernest Dickmanns, a German aerospace researcher. Dr. Dickmanns had, after work with NASA on orbiter re-entry in the 1970’s, returned to academia and started an institute to apply machine vision techniques to create autonomous vehicles. His team retrofitted a large Mercedes-Benz Vario motor coach – chosen because its huge volume could house the necessary computer equipment – and used it as a testbed for self-driving technology. As technology improved, the physical space needed shrank.

It wasn’t until 1986 that nations began to take autonomous vehicles seriously. That year, the intergovernmental research center EUREKA launched the “PROgraMme for a European Traffic of Highest Efficiency and Unprecedented Safety” (PROMETHEUS), a well-funded major initiative to improve the physical architecture of European roadways to improve safety. Their original intent was to deploy guide cables in European highways – but Dr. Dickmanns changed their minds with VaMoRs, a 5-ton Mercedes van that could truly drive itself. On a closed test track, it performed reliably, mastering fully autonomous driving at up to 96 km/hr (60 mph).

VaMP-self-driving-car

Dickmanns used his team’s share of the PROMETHEUS funds to develop VAMP and VITA-2. These highly refined third-generation research cars looked like ordinary Mercedes sedans, but could drive autonomously. The capstone test drive along Autoroute 1 was followed the next year by a demonstration run from Munich to Denmark and back. This drive included parts of the Autobahn, where the self-driving car performed safely while handling lane changes and passing other vehicles at speeds of 75 mph (roughly 110 km/h). Their demonstration proved conclusively that the team had solved all the important problems necessary for cars to drive themselves at the equivalent of human skill level on highways.

So if we had highway autopilot functionality in 1994, why did it take twenty years for this technology to appear in commercial vehicles?

There are three main reasons.

Existing manufacturers change slowly: The design-to-production time for automobiles is several years, and was longer in the past. Each new model represents a major project, with a long supply chain of vendors all building custom tooling to make the parts needed to make the subassemblies needed to build the vehicle. If designers incorporate a new feature today, that feature will not reach the showroom for several years.

Regulation and liability slow deployment: Self-driving cars equalled human drivers more than 20 years ago. But automobiles inevitably get into accidents, and accidents cause lawsuits. Automakers chose to iterate on the technology until the autonomous cars’ driving level had reached a point well above human norms. This insured that in any accident involving a human-driven car, the automated vehicle would inevitably be a provably better driver – a superior stance for legal defense. Although a self-driving vehicle can operate perfectly well using just cameras and basic sensors, modern systems also include LIDAR and other advanced sensors, giving them better perception than human drivers.

Auto manufacturers were waiting for the cost to drop: Since the research to develop the algorithms was done long ago, the cost of self-driving control systems is mostly information technology components like sensors and computers. We often think innovations move directly from the lab to the market, but in reality it is not uncommon for them to languish in limbo waiting for the prices of key technologies to fall. Ray Kurzweil‘s career as a futurist began in his attempts, as a professional inventor, to confront this issue systematically. It is unprofitable for an inventor to develop inventions that must wait many years for the components to drop to an acceptable price. To be able to time his inventions better, he began studying what levels of technology would be needed and trying to predict when the technology curve would reach that point.

What has happened in the intervening 20 years since Dickmanns’ Autobahn demo?

PROMETHEUS ended in 1995, but it was followed by a series of European projects. In 1998, the Italian project ARGO built a self-driving car that drove 2,000 km on Italian public roads in real traffic conditions, using two black-and-white cameras and a Pentium PC running Linux. In 2006 – building on its extensive in-house expertise from PROMETHEUS – Volkswagen demoed a self-driving Golf research race car that could out-perform human drivers. And in 2010, an Italian convoy of four autonomous solar-powered automobiles drove from Parma, Italy to Shanghai, China – a three month long 16,000 km trip.

In the United States the defense research arm of the government, known as DARPA, sponsored its first Grand Challenge desert race in 2004. Although it is not obvious, driving autonomously off-road is much more difficult than navigating highways, since a highway is already an artificial, highly structured environment – perfect for robots. Once the self-driving cars proved their mastery of desert terrain, DARPA changed the venue for its Grand Challenge to an urban course. In 2007, contestants in the Urban Challenge demonstrated excellent autonomous driving skills dealing with traffic. Two years later, Google began its famous self-driving car program.

Meanwhile automakers began rolling out limited autopilot technologies – self-parking, lane keeping assist, automated braking, and “advanced highway cruise control.” Most of these technologies allow the automobile to quietly mitigate the driver’s lack of attention or skill. Toyota now ships eight separate vehicle “active safety” systems that rely on autonomous driving technology. Mercedes 2014 S-class sedans offer a “stop and go traffic autopilot” mode in which the vehicle can drive itself.

Almost four decades later, we approach the breakthrough of autonomous vehicles. These cars are quietly taking over – hidden in plain sight – and are now far superior drivers to humans. That should not be surprising, since they mastered highway driving decades ago.

The long early part of the exponential can be long indeed.

 

About the Author:

David RostcheckDavid Rostcheck is a consulting data scientist helping companies tackle challenging problems and develop advanced technology. He can be reached at drostcheck [at] leopardllc.com.

Like this article?

Please help me produce more content:

Donate!

OR

Please subscribe for free weekly updates:

  • MajorCornwallace

    (From my G+ comment) I took the self-driving car course at Udacity taught by Sebastian Thrun (Google self-driving car guy) back, I guess, in early 2013. It was a fascinating class — very enjoyable and even though the concepts are so abstract (probability expectation matching) the entire time it also felt very practical (we were always talking about real-world programming).

    In short, though, it pointed out just how important progress in conceptual analysis is along-side high-tech and distributed data.

  • Ch. Ed. Culpepper III

    Great article. The chief problem with new technology is that it is handled in an
    unsystematic way. Men like Thomas Alva Edison and Ray Kurzweil have tried to systematize the technological advancement. Although their efforts should have inculcated and raised standards for development cycles, their efforts have not been exploited adequately.
    What self-driving vehicle technology needs is a test track – paid for by DARPA,
    multiple countries, multiple industries, multiple corporations and crowdfunding – where the purpose of the track is to define and measure automated driving standards, with an emphasis on pushing the vehicles to the point of failure. For example, setting a standard for
    anti-skidding on slippery surfaces, where the best performance sets the standard. Once the standard is set, the track is adjusted to push the anti-skid performance to failure. We need standards in self-parking, lane keeping, auto-breaking, cruise control, etc.
    When desktop computers were sold on the basis of their performance specifications,
    the performance on those specifications were under great pressure to improve – whether it was clock speed, RAM capacity, ROM capacity, etc. Moore’s Law was and is a technological pressure that has pushed processor performance with relentless demand. Once we stop looking at automated transport in a generic way, but inspect it by specific performance parameters and associated standards, then the technology will take on the “real” character of exponential tech.

  • Dean Pomerleau

    Nice article. I agree completely that exponential technology that appears to burst onto the scene out of nowhere usually has a long development history.

    But you left out an important chapter to the history of autonomous vehicles. In parallel with Earnst Dickmanns in Germany in the late 80s and early 90s, my colleagues and I were developing and testing automated vehicle technology at Carnegie Mellon University. Our vehicles were driving themselves on the highway as early as 1991. In 1995 we even did a trip in which our vehicle steered 98.2% of the trip from Washington, DC to San Diego, California, which we called “No Hands Across America”.

    Here is a retrospective article on those pioneering efforts to celebrate the 20th anniversary of the event:

    https://www.cs.cmu.edu/news/look-ma-no-hands-cmu-vehicle-steered-itself-across-country-20-years-ago

    Its funny – at the time we predicted automated vehicles would eventually be deployed in about 15-20 years. We weren’t too far off!

    Dr. Dean Pomerleau
    Senior Research Scientist, Adjunct
    Carnegie Mellon University Robotics Institute
    Pittsburgh, PA USA

  • That looks like a great class – it’s on my list to take and I’m looking forward to taking it.

  • I did miss this – thanks for bringing it to my attention. We’ll have to do something about that (more below). To tell you the truth, I found it quite surprising how hard it is to dig up history on autonomous driving cars. It’s not nearly as well curated as I expected it to be. Since much of the work was done pre-web or just at the beginning of the web, it tends to not be in search engines. In fact I got into the topic when doing research on Deep Learning and reading a paper by Jugen Schmidhuber pointing out that the dense citation network of the “Deep Learning Conspiracy” (LeCan/Hinton/Bengio) – which defines the tech press narrative – tends to completely ignore an entire major batch of development from Europe. While looking into that, I stumbled across Dickmanns work via a mention from Schmidhuber. As I learned about it, I mentioned it to Nikola, with the sense of “Can you believe this? We had viable polished autonomous cars driving around in 1994 and it’s not well known at all,” and he suggested I dig into it and write an article for Singularity WebLog about that.

    I did try to find more about the U.S. efforts but found the available information to be thin and scattered. I came across ALVIN (early DARPA effort), but it didn’t seem to connect to anything else. One of the sources I used was an exhibit from the Computer History Museum here: http://www.computerhistory.org/atchm/where-to-a-history-of-autonomous-vehicles/ and another was a trajectory through Wikipedia – which were also silent on the CMU work.

    Your information does answer a significant loose end I had left: why was CMU such a strong contender in the DARPA Grand Challenge? Where did that knowledge come from? I made some efforts to find out, but the article was getting long and I was slow in getting it to Nikola so I wrapped it and went with it. But there might be a good opportunity for a follow-up here, if you would be willing to help out as a primary source, since this space is so sparsely documented.

    And one thing we really need: more public domain photos! There is a real dearth of them. If you have any photos of key vehicles, team members, and other milestones that you can put into the public domain, that will help considerably in getting the story of that branch of research better known.

    Thanks again for bringing this to everyone’s attention!

  • That’s a good idea. One related thought that I had when doing this writing was, given the rise of computing in cars and the attendant issues with security through obscurity, will we eventually see the rise of open-source cars (and autonomous cars)? That would give an opportunity for systems to be benchmarked against each other, which would help drive performance through competition. And such a thing might well serve as another route to a crowd funded test center for autonomous automotive technology. Another route is an industry-wide consortium for testing, perhaps evolving out of existing shared industry-wide resources – but as we see in the article, the automative industry has a great deal of inertia.

  • Deborah Henderson

    Surprisingly, I also worked at the Field Robotics Center in the late eighties and early nineties. I might have some photos lying around. The Nav Lab wasn’t my project, but I probably still have pictures.

  • Well, hi! Apparently I should have asked people I know 🙂 If you have any you can find, that could definitely help.

  • Dean Pomerleau

    David,

    I’d love to help you out with your research on the early days of automated vehicle research in the US. Here is a good recent article (with historical pictures) written by one of my colleagues on the project, Dr. Todd Jochem:

    http://jalopnik.com/they-drove-cross-country-in-an-autonomous-minivan-witho-1696330141

    Its got a lot of helpful links to other resources on our research and development in the early 90’s, including ALVINN – which BTW was my baby!

    Here is a paper I and several other colleagues (Chris Urmson, Todd Jochem, Chuck Thorpe and Dave Duggins) authored on the history of self-driving car R&D at CMU from the mid 90s until the time of the DARPA Urban Grand Challenge in 2007:

    http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4640916

    It is behind a paywall, but I’ve uploaded the paper so you (and others) can access it as well:

    https://www.academia.edu/16837347/From_Automated_Highways_to_Urban_Challenges

    You’ll notice the name Chris Urmson as one of the authors. Chris should be familiar, since he leads Google’s self-driving car program today. So there is definitely a direct link between the early efforts in self-driving cars that we did at CMU in the late 80’s and early 90’s all the way to today’s automated Google cars.

    Let me know if you would like more info, or would like to speak directly with me about the project. I’m happy to help!

    –Dean

Over 3,000 super smart people have subscribed to my newsletter: