2016 is looking set to be a big year for self driving cars. CES 2016 has just flown by which has shown off some technology for cars in the future. Actually, saying this, the future is here already. The Audi A7 can drive itself on motorways and change lanes – the Tesla Model S can do the same too. So, I guess when we are talking about the self driving cars of the future, we mean the true self driving experience. Getting into a car and not having to touch anything at all – just letting the car drive you to your destination, that’s self driving. However, as great as this does sound, there is still one fundamental flaw that will always be hovering over self driving cars just like a grey thunderstorm cloud…

Driving into a tunnel Conundrum

The main overshadowing problem with self driving cars is, what I like to call it, the ‘driving into the tunnel conundrum. The situation is that you are on the motorway going about 70mph and you are about to head into a tunnel inside your self driving car. The chances are you have your feet up and maybe catching up on some Netflix! Who knows. Either way, you are not driving the car and are not intending to have control of the car any time soon.

All of a sudden, just before entering the tunnel, a small child runs onto the motorway straight into your lane. It is too late to brake so there are two possibilities from this situation:

1. The self driving car swerves around the child to save the child’s life. But, in doing so, the self driving car hits the side of the tunnel at 70mph and kills everyone on board from the huge impact.
2. The self driving car tries to slow down but kills the child in trying to brake knowing that swerving will cause those on board to die from the impact of hitting the tunnel wall.

Let’s be honest, both outcomes are tragic and horrible. However, when you are about to automate a feature such as driving, these limits need to be tested.

The problem occurs that nobody will want self driving cars on the road if they know that if their child runs out into the road, the likelihood is that the self driving car will choose to kill the child rather than the occupants inside the car – pedestrians will get very scared by this. The other problem is that if the self driving cars chose to smash into the tunnel wall to save the child’s life, who is likely to buy a car where the life of those outside the car is valued more than the life of those inside it?

From this, it really does seem a catch-22 situation – you are dammed if you do and dammed if you don’t. For this reason, it will be very interesting to see how those manufacturing and creating software for self driving cars overcomes this scenario…