Wednesday, 2 October 2013

Why Fully Self-Driving Isn't the Automaker's Goal

A number of automakers have been making statements recently around the subject of automated (autonomous, self-driving, driverless) cars.  The general message seems to be: 'We are going to add safety features to our car that will automate most or all of the driving task, but the driver will always need to be in the loop and of course the pleasure of driving is so great that there will always be times when the driver will just take over.'
For example:
These attitudes in some ways highlight the automakers quandary. They have a tried and tested 130 year old business model of incremental improvements where they seek to sell every single one of us at least one vehicle. They sell speed, power, luxury, connectivity, versatility, safety (except some of the previous sales points are in direct contradiction of safety?...) and even efficiency, as the younger generations are so much more aware of 'green' issues.

Such a business model is great for the shareholders of these automakers - but considering most vehicles stand around idle for 90% of the time, this can be considered a serious waste of the earth's finite resources. Also consider that human error is a factor in some 93% of crashes - we have a terrible conundrum where there is an entire industry, including aftermarket parts, that relies on crashes occurring.

Whereas a truly self-driving, NHTSA Level 4 automated vehicle, one that is capable of driving unmanned, challenges (if not disrupts) this business model. Instead it inevitably leads to the development of automated shared mobility fleets - especially from the business models of taxi, car-rental, car-share, ride-share and P2P companies.

We know from studies that 1 shared vehicle can typically take 9 to 13 private vehicles off the road (Shaheen, UC Berkeley). Studies around the shared automated fleets, or 'aTaxis' suggests that 1 fully automated Level 4 car could take at least 2.5 private vehicles off the road during peak hours (Kornhauser, Princeton) - and maybe more.

With these aTaxis the average person could relinquish ownership of their private vehicle and hire the right size vehicle for their commute to work, family time at the weekend, shifting goods etc. - and save 40% of annual transportation outlay in the process (as extrapolated from 'Transforming personal mobility - Earth Institute, Columbia University).

Therefore with and aTaxis we would find that the automakers sell many more cars to the fleets and less to private owners. Which means that their 130 year old business model no longer looks so robust. Fleet owners will want very different characteristics in their cars and they will be very aggressive on beating down prices.

So comments like the ones automakers make in these articles about 'keeping the driver in the loop' and that 'drivers will always want to drive' etc. don't really stack up when you consider the likely new business models and the societal benefits that Level 4 automated vehicles can bring.

Also note that the 2020 date for this technology from the automakers is not for Level 4 automated technology - but Level 3 (tending towards Level 4), where the driver is required to take over the driving task when necessary - because that preserves their business model - 1 driver, 1 car (or more).

Whereas Google have explicitly stated that their aspiration is to go for Level 4 if they can manage it, which would mean bypassing Level 3 altogether. On the Friday following this presentation by Ron Medford of Google there was a public meeting regarding autonomous vehicles in California.  There I took the chance to ask the panel, about when they thought fully self-driving cars would be available. Anthony Levandowski of Google replied that they stand by Sergey Brin's statement at the California autonomous vehicle bill signing, which was where he intimated that they are aiming for about five years, i.e. 2017.

So when you read what the automakers are saying on the subject of driverless cars, please consider that there is another side to the story and the business models and agendas lie behind much of what is said.

2017 is the date that we should be preparing for. To not do so risks all sorts of avoidable societal and economic collateral damage.