Skip to content

The cost of no driver

Industry insiders have been discussing the idea of self-driving cars since the 1980s but the technology at the time was not sufficient to make the idea a reality. It took the intervention of a whole new industry, the tech companies in Silicon Valley, to make it happen.

Industry insiders have been discussing the idea of self-driving cars since the 1980s but the technology at the time was not sufficient to make the idea a reality. It took the intervention of a whole new industry, the tech companies in Silicon Valley, to make it happen. Now Google has taken over the mantle and its extraordinary driverless car is the talk of the industry but does it really have to be so expensive? Google announced that it had built such a car in 2010. In true Google fashion the California company threw huge amounts of money and engineering talent at the project; because they could.

The result was successful. Since then Google's self-drive car has covered more than 700,000km without mishap. Google is delighted and is happily informing everyone that their car is the car of the future. What they are less vocal about is the money it costs. Google's car achieves its impressive performance by processing a gigabyte of data per second. That is a lot of processing power and it requires expensive computer equipment to do it. It also requires a laser range finder which itself costs $70,000. The technology cost for each car is around $150,000, which many limit uptake somewhat.

Google tells us of course that these costs will come down as production scales up and the tech becomes more mature. The car industry is less sure and executives quietly doubt that Google's technology will ever become cheap enough for a mass market product. That does not leave the idea dead in the water though. Other companies across the world are working to develop systems for driverless cars that are much cheaper. Right here in the UK for example, boffins at Oxford University are developing a system to cope with British weather conditions that are unheard of in California.

This system works by using a laser scanner to recognise where it is by comparing its surroundings with stored data. Google's system on the other hand uses GPS, laser range finding and mapping to work out where it is and what route it should take. The interesting thing about the Oxford University approach is that it is claimed that it could be retro-fitted to existing cars and, at some point, cost as little as £100 to fit. Clearly that makes it a much cheaper and more adoptable solution than the Google car. But will British thrift win out over the lavish systems of the big boys from Silicon Valley? We will have to wait and see.
 

Posted by on

Back to June 2013

Back to top