Uber Self-Driving Car Sounds Cool, But Its Future Rests on Humans
Cars embedded with chips respond to signals they receive. 
Cars embedded with chips respond to signals they receive. (Photo: iStock)

Uber Self-Driving Car Sounds Cool, But Its Future Rests on Humans

This week’s incident, which involves an Uber self-driving car killing a woman in Arizona, has raked up the debate on safety in the autonomous driving ecosystem. Till now, most of these accidents involved two cars without any human casualty, but this time it is different, and Uber and the others know it’s serious.

The use of radars and cameras ensure that such autonomous cars can navigate public roads. However, technology can never match the human eye. Various research experts have pointed out that self-driven cars will result in fewer accidents, but the Arizona incident clearly doesn’t inspire any confidence.

In order to put this issue to rest, auto manufacturers should be heading back to the drawing board for autonomous cars.

Also Read : Uber Not at Fault in Self-Driving Car Accident, Says Police

As you might know, self-driving cars are heavily tuned to work under certain conditions and road standards that aren’t available all the time.

The Tesla Model S incident from 2017, where it crashed into a trailer is a prime example of that. According to the investigation, the victim did not adhere to the car’s instruction of taking control of the car while he was travelling.

The self-driving cars heavily operate under the guidance of hardware and software that are embedded into the system. However, you can’t discount the element of human control that is required to prevent any mechanical mishaps.

How Waymo’s self-driving cars detect ongoing subjects. 
How Waymo’s self-driving cars detect ongoing subjects. 
(Photo: YouTube/Waymo)

Currently, autonomous cars operate at five levels, which highlight its capability and also entrust the driver to play a role while the car is in motion. Take a look at how it works out:

  1. Level Zero - Cars are completely controlled by a driver
  2. Level One - Cars assist with some functions, but major aspects covered by driver
  3. Level Two - Car controls the steering and acceleration (most followed at the moment), but sudden braking requires driver intervention
  4. Level Three - Conditional automation embarks on the car controlling most driving aspects via sensors like LiDar
  5. Level Four - Cars can handle steering, braking, change lanes and use signals also.
  6. Level Five - Fully automated vehicles, with no pedals and even steering wheel fitted on cars at this stage.

Also Read : Self-driving vehicles cause higher stress levels to passengers

Headed to a Future of Automated Driving

Researchers have also stated that the sensors on-board these cars end up being less reliable if you’re running under bright sunlight, which again, is bound to happen in India.

With such cases in mind, it is evident that the use of self-driving cars, at least for now, should be heavily dependent on human intervention. They will have to remain semi-autonomous at best. We’ll have to wait for Uber to come out with a full explanation on the cause of the accident to give them a direction to work on.

Also Read : Google’s Self-Driving Car Project Gets a New Name: Waymo

Pedestrians, cars and other moving subjects need close-eye tracking. 
Pedestrians, cars and other moving subjects need close-eye tracking. 
(Photo: YouTube/Waymo)

However, with over 59 cases involving autonomous cars reported since 2017, companies need to realise that competition can take a backseat. It doesn’t really matter if one brand has managed to beat the pack by offering better services.

Instead, they should only roll out self-driving cars after tests in most extreme conditions, collaborating if necessary, for the best of technology.

Other than the US, where various states have specific self-driving guidelines, countries like China, Japan and even South Korea ensure that a driver is always made available in the car, which in some cases negates the need to issue licenses.

The biggest weakness for such technology is its lack of preparation for surprise elements. Roads can be full of surprises, and as it turned out for the victim, extremely dangerous as well.

How self-driving cars look from the inside. 
How self-driving cars look from the inside. 
(Photo: YouTube/Waymo)

While such accidents have only hit Tesla and Uber until now, chances of Waymo or other self-driving manufacturers falling prey will be hard to dismiss after this week’s episode. Toyota has already stopped the testing of its self-driving cars, and others are likely to follow suit.

This also raises the question of self-driving cars succeeding on our shores. In a country like India, where you have pedestrians, two-wheelers, cyclists and even animals, cars like these won’t survive without an incident even for one day. Imagine one of these Uber cars coming in close contact with a pedestrian or even an animal. These machines may not have the reflexes to handle the Indian driving conditions.

(The Quint has given up the use of plastic plates and spoons. This Earth Hour, what will you #GiveUp to save the planet? Use the hashtag #GiveUp and tag @TheQuint to tell us.)

(The Quint is now available on Telegram. For handpicked stories every day, subscribe to us on Telegram)

Follow our Tech News section for more stories.

    Also Watch