Introduction

Advances in modern automobile technology have completely changed how we commute. Although we’ve been unable to recreate lightsabers, we’ve built something equally awesome: driverless cars.

How They Work

The way in which driverless cars reason about their surroundings is not very different from us humans. Using radar, lidar, cameras as well as some peripheral technologies, driverless cars are able to sense not only the presence of objects but also their velocity and their size. This raw data is then processed to calculate trajectories and other similar factors. The car responds to this data by turning, speeding up and stopping. This cycle is repeated extremely fast, thousands of times a second, allowing the car to achieve response times much, much faster than humans.

This isn’t very different from how we humans process information. Our senses perceive stimuli, our brain processes them, and we decide how to respond.

The reason why driverless cars are groundbreaking is that they’re autonomous, not automatic. A driverless car doesn’t need the installation of any special equipment on roads, which means we don’t need to spend money on installing additional infrastructure on roads. You just need the car.

* Lidar is a technology similar to radar, where you bounce some sort of wave off the environment and see how the environment responds. In the case of radar, it’s a radio wave, while LIDAR uses a laser.

The Big Players

There are a few big players in the driverless game, but the industry is lead by Tesla Motors and Google.

Tesla rolled out its semi-autonomous Autopilot feature to most of its cars in September 2014. It’s not completely autonomous, i.e. there still needs to be someone behind the wheel to take control of the system during any difficulties. Tesla Motors was one of the first companies to market driverless technologies to ordinary consumers, starting nearly 5 years ago, and remains a leader in the commercial driverless car industry. Unlike most other driverless cars, Tesla automobiles have a consolidated body. External components aren’t required since the sensors are built into the car itself.

Google isn’t a car maker. Its driverless car project is part of a larger effort to aid the transition to electric cars for cleaner energy and its cars have driven nearly 3 million kilometres. There were only a few accidents, and all of them were attributed to human error.

A Driverless World

Elon Musk, the founder and CEO of Tesla Motors has stated that he believes that eventually, human driving will be outlawed because of its dangers.

Although it is difficult to entirely eliminate traffic, traffic is still an area where driverless cars can help. A network of driverless cars could intelligently avoid situations that increase the

chances of having a traffic jam. Driverless cars will also open up a lot of possibilities for refueling, since driverless cars could just go refill their tank / battery themselves, without any human intervention.

Where Do The Jobs Go?

In New York City, a taxi medallion is a license required to pick up people who hail taxis on the street. The cost of a taxi medallion was 1 million dollars a year ago, but with the introduction of ridesharing services like Uber and Lyft, the price has fallen to just half a million dollars. Because of this, taxi drivers are unable to sell their medallions and join Uber or Lyft since they would be making a loss. Driverless cars would be another nail in the coffin of the nearly 42,000 taxi drivers of New York City, driving the costs of taxi medallions further down and leaving taxi drivers stranded.

Driverless cars would poach jobs from taxi drivers, Uber and Lyft drivers, truckers and bus drivers, causing an unemployment crisis. 0.3% of the US GDP comes from these individuals, who earn a mean income of 42,000 dollars and account for 67 billion dollars of income.

We could see a crash reduction rate of up to 90%, leaving 445,000 automobile repair technicians without jobs since people would visit their shops less often.

The Trolley Problem: A Thought Experiment

There is a runaway trolley hurtling down the railway tracks. If the trolley is allowed to move on the tracks unimpeded, it will kill 5 people. You stand near a lever which can divert the trolley to a different track, but you notice that there is a person on the other track as well. Would you divert the trolley, killing one person but having to bear the responsibility for his death, or let the trolley continue on, killing five?

Another variant of the trolley problem is one where you’re standing on a bridge, above the railway track. Rather than pulling a lever to divert the trolley, you may push a fat man off the bridge and onto the path of the trolley in order to stop it. Would you push the man off the bridge or let the trolley continue on and kill the 5 people?

The relevance of the trolley problem in the world of driverless cars is in a situation where the car has to either run over a pedestrian or swerve and cause the death of the car’s occupants. It is a difficult problem and doesn’t really have any solutions. If you want to offer your opinion on dilemmas like the trolley problem, then you can take MIT’s Moral Machine survey at moralmachine.mit.edu. It’s very interesting and pretty thought-provoking and I strongly recommend you take it.

Liability

The trolley problem leads us to an interesting legal problem: if the car hits a pedestrian, is involved in an accident that kills the occupants, or crashes, then who is liable? Is the manufacturer or are the occupants of the car? Pinning liability on the manufacturers is problematic; making manufacturers liable for accidents demotivates them from manufacturing the driverless cars in the first place, and ultimately stifles innovation. On the other hand, it doesn’t make sense for the occupants of a completely autonomous car to be liable for accidents.

Volvo, which hopes to roll out its driverless IntelliSafe Autopilot system by 2020, says that it will take responsibility and pay the damages for accidents caused by its driverless cars. Erik Coelingh, its senior technical leader for safety and driver support technologies, said that the Volvo’s IntelliSafe Autopilot system will eventually include so many redundant and backup systems that a human would never have to intervene, and could not be at fault. It’s hoped that other car makers will follow suit, but liability, in this case, is a complex legal problem and it will take several lawsuits before it is fully understood how responsibility should be attributed in the event of an accident.

Uber’s Driverless Cars: A Case Study

The city of Pittsburgh is home to bridges (446 of them!), tricky intersections, hilly areas, diverse climatic conditions and a large variation in terrain. It’s also home to Carnegie Mellon University, which has one of the most advanced robotics and machine learning programs in the world. This made it a perfect location for Uber to test out its fleet of self-driving cars, which it rolled out in September 2016.

The fleet right now consists of modified Ford Focus cars. The cars have 20 cameras, seven lasers, GPS and radar equipment, and a liquid-cooled computer in the trunk. If you’re ever in Pittsburgh, you can recognize these cars from the spinning lidar unit on the roofs and the Uber logo across their sides.

The cars are requested in the usual way, by ordering an Uber ride from the app. The app randomly pairs you up with a driverless car, and for the time being, rides in these driverless cars are free.

Since the cars are currently in testing and are not completely ready, they still require an Uber technician to be at the wheel and ready to take over at any moment in case of difficulties. There’s usually an Uber engineer sitting next to the technician, noting down any unusual occurrences and ensuring everything is going smoothly.

The cars are currently not able to navigate bridges which is ironic since Pittsburgh is often referred to as ‘the City of Bridges’, in reference to the city’s whopping 446 bridges. Bridges pose a unique challenge since they lack contextual cues like landmarks and buildings.

During a visit to Pittsburgh, I had the unique experience of seeing several of Uber’s driverless cars for myself. They’re easy to spot, and you’ll see several if you’re walking around the city.

What About India?

While European countries and the US are wrapped up in the testing of driverless cars, we have to wonder: what about India? Will India ever be ready for driverless cars? At the moment, driverless cars require well paved roads with clear markings. India lacks the infrastructure for these cars to run, with only 16% of all roads being paved. In a country where roads lack basic traffic signs, a car which reacts by reading these signs and markings would run into problems pretty soon. Sudden obstacles like cows and other animals also pose serious problems.

To shine a bit more light on how an Indian techie would view driverless cars, we were lucky enough to have Rajiv Mangla, the CTO of Snapdeal share his opinion. We asked him about the scope of driverless cars in the future of India. He believes that it will leave many unemployed but will make for well-planned cities. Daily commutes will be much more predictable. However, he feels that India isn’t going to see commercial driverless cars in the near future unless the technology in the cars and the quality of roads is improved.

This doesn’t mean that Indians can’t make driverless cars. A few Bangalore based engineers have already retrofitted a Tata Nano with sensors and cameras to turn it into an autonomous vehicle. Roshy John, the person behind this project, began his work in 2011. Although the team is yet to take the car for a test drive, it means the future for Indian autonomous vehicles may not be too far away.

What The Future Holds

It’s hard to say what the future holds for driverless cars since they face a number of hurdles. Many people are still reluctant to sit in them out of safety concerns. It’s up to companies like Uber, Tesla, and Google to show people that driverless cars are the future, and that the future is now.

Written by Kabir Goel. Section on India written by Sarah Randhawa.