Advertisement

Blog

Propagation of the Autonomous vehicle technology and its epicentre

Almost anything that moves can become an autonomous vehicle. Tiny autonomous shopping carts are on one end of the spectrum and gigantic autonomous long-haul trucks can be considered the other extreme, if we were to think of only vehicles on road. The testing of self-driving cars by Google (now Waymo) began way back in 2009 on city roads. This was indicative of the technology first hitting the consumer/passenger space. This thought, however, met some policy resistance and consumer reluctance, lately.

The question that we are faced with now is which industry autonomous car technology is going to hit first. Keeping the need in mind, factors such as rising labour cost and unavailability of drivers for trucking industry allude that there is a potential opportunity in this space for autonomous vehicle technology. Data validating the increasing labour costs comes from, for instance, China, where the cost is likely to double from 2013 to 20191 . Also, since autonomous trucks will imply more of highway driving, the driving environment is favourable. It is less controlled than in mining and agriculture but much more controlled & predictable as compared to in-city autonomous driving. As per the American Trucking Association the trucking industry today employees close to 3.5 million truck drivers2 . The cost savings in trucking industry with the advent of autonomous trucks is attributed to elimination of wages (approximately 40% of the operating costs) and the rest to factors like reduction in insurance sums.

The Logistics and Delivery space seems to be the one seeing major upheaval. Waymo is going to have its trucks deliver freight for Google data centres in Atlanta. This is apparently their pilot program to evaluate the effectiveness of integrating Waymo technology with logistics. Amazon’s latest patent of a ground-based autonomous delivery vehicle further reinforces the reports by analysts mentioning that the last mile delivery space will be worth 80 billion USD by 2025. Autonomous Robotic vehicle start-ups like Nuro are entering the last mile delivery space through partnerships with retail giants like Kroger. Thought leaders from the likes of FedEx, UPS, DHL are already talking about the technology. Delivery and logistics space thus seems to be the one seeing hints of the technology at the first and the last nodes, and in the next decade one can expect end-to-end autonomous delivery.

Factors driving market adoption

Accuracy, cost and social acceptance of the autonomous vehicle technology are three important decisive factors. The percolation of autonomous technology into the trucking and logistics space will help it scale up owing to cost reductions. Cost and hence price, is the first barrier for adoption of autonomous technology in consumer space. When the price is low enough for a citizen to be able to afford it, say, on a Lyft trip, the technology will penetrate into the car sharing space. As per reports, in Los Angeles $1.55 is the cost per mile of an Uber while the cost of ownership of cars comes to about 54 cents a mile. As per some predictions, autonomous taxi costs will decrease from around 85¢ per mile in 2018 to about 35¢ per mile by 2035. This will be lower than the cost of ownership of cars today, justifying the economics of owing an autonomous car. A self-driving car needs to be able to beat those numbers.

The second barrier, is the social aspect, how comfortable people are with using autonomous cars. The recent incident in Arizona resulting into the death of a 49-year-old woman has, however, been a derailer for autonomous technology. Where 67% of Americans felt less safe with autonomous technology until a year back, now the number has risen to 73%3 . Displaying the efficacy of this technology in trucking space with social experiments by Waymo in different cities, like Phoenix, will help reduce the psychological and social barrier eventually. As adoption increases the cost will reduce further which will in turn drive more adoption, and this will go on. Also, as the usage of autonomous vehicles increases, since it’s a learning system, the technology will become more accurate. Increasing accuracy will lead to faster reduction in social barrier.

Passenger mobility space, alignment of the Automotive players

In the year, 2025 close to 40,000 miles in US will be autonomous shared driving covered. In 2030 people can be expected to “own” self-driving cars. The automotive companies have started positioning themselves in the car sharing space to ride this wave from the very beginning. Automotive giant Daimler AG has car2go, BMW has DriveNow, AUDI has AUDI on demand, GM has Maven indicating that shared mobility along with multi modal transport is the way to go. In the multi-modal space, Daimler AG with its Moovel app has introduced some direction on what is the need of our times in terms of transportation. The industry is polarizing towards technology start-ups and disruptors. GM acquired Cruise Automation, which is amongst the spear heading companies in the space. Fiat Chrysler, Lexus, Jaguar Land Rover are working with Waymo. Other leading automotive companies like Tesla are in news for building their own Autonomous vehicle solutions.

Conclusion:

Going forward, with autonomous vehicle technology, the boundaries between different areas of application of vehicles may dissolve. An autonomous taxi, not driving at some time, will end up completing a food delivery and then pick another passenger and drop him somewhere. Autonomous technology will probably enter all nodes of the end-to-end transportation system, autonomous shuttles, trams, trains, bikes, taxis etc. all embedded in a single platform in the next decade or so. Accuracy will drive adoption and more adoption will keep improving the accuracy. The cost of ownership of autonomous cars will decrease as the technology scales up. The whole point then really is which industries are going to be the earliest adopters & when will we see this technology creating a positive impact in the public safety space.

The author would like to acknowledge Partha Mukherjee and Neil Gomes 

“The views expressed in this article are mine and my employer does not subscribe to the substance or veracity of my views.”

References

1 McKinsey & Co.: Gearing up for growth, Future perspectives on the global truck market

2 American Trucking Associations: Reports, Trends, and Statistics

3 Valuations Sizzle, Canada Fizzles and Autonomous Vehicles get Rolling Again

7 comments on “Propagation of the Autonomous vehicle technology and its epicentre

  1. herbissimus
    August 26, 2018

    Sweet, but will I or the other thousands, perhaps millions of possible personal vehicle purchasers really want to own something that spends most of its life parked in a garage or on the street when one that drives itself could show up at your place within 10 minutes of being summoned from one's smartphone. Yes , the vain, early adopters, and the wealthy might want their own, but the rest of us, maybe 99% of us won't because of the ancillary cost of your vehicle sitting dormant for over 90% of the day, not to mention the cost of a garage.

  2. GarimaJain
    August 27, 2018

    Hi,

    Couple of comments to that. With AV, a tendency of co-owning of cars is also developing. This means more than one family/person will own a car together so that the AV doesnt waste its time in a garage. Secondly, I am sure before AVs are owned they will be used by the very companies whose cars you summon today. Lyft, Uber etc. have been in news for their AV efforts. This implies that may be aroud 5 years from now when you do call for a car using your smartphone an AV will stop at your doorstep 🙂

  3. herbissimus
    August 27, 2018

    After seeing a young man in the silicon valley take a honda and use an android phone to automate it to level 2.? I got the flash that the retrofit market might well convert many of the present day autos into self driving ones. We could well see 1950's Buicks without drivers.

    Search George Hotz.

  4. GarimaJain
    August 28, 2018

    Thanks for that input. But the struggle amongst the major automotive companies and the tech disruptors like Waymo, Uber is not level 2 automation, but level 4&5. Level 2 automation is simple conditional automation (if I am not wrong) and doesnt need AI. And level 2 automation vehicles have been in use I think for the past couple of decades.

  5. herbissimus
    August 31, 2018

    Its been awhile since  I read about Goerge's exploits. what he succeeded in doing was making the car self driving on the freeway. keeping to its lane and avoiding the other autos at freeway speeds. Ithink the jury is still out over what forms of computation will make the best, most cost effective strategy for achieving  self driving. all the fears that exist about networked security could well mean that the decision making will all be local to the vehicle, and there'll be no communication with the external environment, as an example, the auto will make decisions locally, not over any network and any software changes will be with a wired connection to the parent of the vehicle.

  6. GarimaJain
    September 6, 2018

    Yes! you are absolutely right, the decision making will reside locally in the car. Some inputs like HD maps will reach the car in batches. Your point on OTA vs wired updates is duly noted, I am yet to form an opinion on it. I will have to understand more from the Tesla example. Thank you!

  7. UdyRegan
    October 5, 2018

    The autonomous industry will never form a complete circle in terms of the progress that it is making currently. There will still be faults that need to be considered before they can be adjusted accordingly to prevent major problems in the future. There was a case of a pedestrian being struck and killed by an autonomous Uber and this is a clear example to why this industry still gives people all the reasons to frown. There still has to be some sort of human intervention which defeats the purpose of this industry even more.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.