Autonomous Cars The Eventual Impact
A compilation of thoughts and questions on the autonomous era.
The Automobile era began in the 20th century, although for the first few decades it was considered that owning cars equaled to opulence, it soon became more widespread to a point now where it’s become almost a necessity, one of the first things that any household seeks to buy in todays era. Not having a car is not as such looked down upon, but is something families strive to buy as a first thing to possess in their lives. What has this resulted into? We now live in a world struggling with huge number of vehicles and the traffic which results thereof.
This huge amount of vehicles has led to kilometers long traffic situations. An average commuter in a metropolitan city almost spends about ~ 50 hours in traffic. These vehicles result in carbon emissions due to their usage of carbon fuel. But I’m not really here to sum up the various effects and consequences of automobiles, I’m here to provide my thoughts on what we might be moving towards. A pure driverless car era ushers us towards a reimagination of transportation as we know it. One of the first things that come to my mind is this article I read on MIT tech review, it spoke about how autonomous cars might rather chose to prefer to not park at all or park at some distant place to avoid the billing in the local parking spaces, which would render these places useless. But then the effect this would have is even a small fleet of 2000 cars would drive the traffic in san francisco down to 2miles per hour. One thing to take away from here is the transition from purely driven cars to self-driven cars will be very difficult, since this coexistence would imply dynamics not usually seen because when people drive they expect other drivers to drive in a similar way because of some unspoken driving rules usually conformed to. When you introduce a self driving car in such a dynamic, not only is there the danger of non-conformance but also doing something entirely opposite.
One thing I’m extremely excited about in a pure autonomous vehicle era is the probable disappearance of traffic. The current traffic handling mechanisms including signals and other stuff is in place because of the absence of omniscience at intersections and roads in general. If you had cars driving in a swarm or in an area communicating with each other, they could be designed to avoid each other unless the area of crossing is too narrow. At intersections imagine no signals and the swarm of cars from each direction don’t ever stop because the knowledge of direction and travel among these cars allows for extremely smooth handling. This would reduce travel time, allow better control, maximize efficiency of fuel and many other details like so.
But its not an entirely easy way towards there, its not that the grass is greener on the other side because even to get there we need to consider several things which are arise from the introduction of such cars. One of them is attacks on these models by which the cars are being driven. Autonomous driving is mostly done by capturing live feed of the road and deciding what to do given what has been captured, but given an adversary who could hack this feed or introduce some sort of occlusion in the environment which leads the car to either entirely miss the object/place/person or mistake it for something else entirely might lead to very deadly accidents. Imagine an occlusion when introduced makes the car to think that the thing in front of it is an empty road while people are crossing by, if it were a person then he would probably stop it directly. The more autonomous cars become digital the more you need to think about adversaries not just to the model itself but also taking control of the car remotely which then makes it a cyber security problem. One thing I’ve not yet spoken about the second and third order effects that the introduction of these cars bring. In places like USA, autonomous trucks could eventually displace about 5 million jobs. Its not just that, the second order effect which is really noticeable is motels, and roadside bars and eateries which have sprouted across highways for drivers who use these facilities to get rest on a long trip. In an autonomous setting, although not immediate and not entirely but most of these places will probably be shut down or remodel themselves to appeal to people who still might want to drive on their own. There are a lot of other effects which might be introduced and I will update them as and when they appear to me.
One of the final things I would like to return to would be the thing i started with i.e how cars became a necessary possession for human beings, but I think that this idea of possession must morph into something else in an autonomous setting because you would have cars going from one place to another all the time and you could have some decentralized application which lets you get ride these cars in with other people on the same trip.
Although I might have set a grim tone towards the latter half of the article. It’s a far fetched thought to think about those effects now. A lot of what i covered and thought about make sense in developed countries like China and USA since they already have pilot program in place and are looking at a slow yet gradual shift to autonomy, but in the case of lesser developed countries which are still in the second or first stage of autonomous vehicles it might take a few decades or even half a century to start considering these issues since they either don’t have the infrastructure to support these vehicles or might be too skeptical. In India and many such places in the world, there are a lot of people who depend on driving for their livelihoods and the introduction of autonomy might leave them jobless. Not only that, but there a lot of people who like driving on their own from one place to another and although autonomous cars will be introduced people wouldn’t necessarily take a liking to them. When you ride a car from one place to another you trust the driver to do his job, So the question to ask here really is, even though statistically speaking the introduction of these cars will reduce accidents and deaths, are we really at a place where we can trust machines with our lives ?
The MIT Tech review article I spoke of in the article can be found here.
Thank you for reading the article.