REGULATIONS FOR SELF-DRIVING CARS

Ed. Note.: This post, by Vishal Rakhecha, is a part of the NALSAR Tech Law Forum Editorial Test 2016.

Self-driving cars have for long been a thing of sci-fi, but now with companies like Uber, Google, Tesla, Mercedes, Audi and so many more conducting research in this field they don’t seem as unrealistic. Self- driving cars are vehicles which do not require human supervision, with autonomy of varying degrees. Such technology is already present -to a limited extent – in the form of cruise control, parking assist, etc. The creation of such technology would inevitably require a sound system of rules and regulations. These laws among other things must be capable of setting a set of standards for the companies, securing the physical safety and protecting the privacy of the end user. Presently Motor Vehicles Act, 1988 and Central Motor Vehicles Rules, 1989 are the only rules related to automobiles. These laws are inadequate in terms of their application to autonomous cars. This article deals with the changes in the law which may be required to deal with challenges which this new technology may present. These modifications will be essential to ensure the protection of all stakeholders when these contraptions do come on Indian streets. This article will deal with regulations of self-driving cars of level 3 and 4.

The National Highway Traffic Safety Administration in the USA has segmented the autonomous cars in 5 levels.  For the purpose of this article we need to understand what levels 3 and 4 are. In level 3 the car has a high level of automation and does not require the driver to constantly monitor the roadway during the trip with brief periods where the driver control is necessary. The cars in level 4 are completely autonomous and are capable of performing all safety-critical functions. The user is only required to put in the destination and navigation details.

There have to be certain specifications and features based on best practices of the industry which each car must have to get clearance from the government authorities. Each car must have, other than the very basic features required for it to be autonomous, a steering wheel, pedals and an overriding mechanism to give control to the human operator at any time. The tests would include capacity to sense obstructions in front of it, to interpret and adhere to traffic signs and understand signals given by the other drivers on the road mechanical or by hand, ability to follow instructions as provided by its operator, come to sudden stops and increase speed, change lanes, ability to identify smaller objects like children, cycles, pedestrians, etc.

As mentioned earlier both CMVR and MVA have provisions to deal with vehicles requiring constant human supervision. In these laws the probable consequential placement of liability in cases of mishaps is related to only these types of vehicles. Therefore to give these cars a legal backing to actually be able to operate autonomously on the street it is important to include them in the ambit of the definition of the word ‘driver’. A possible version of this could be- ‘a machine capable of manoeuvring itself and that which has passed a driving test.’

The Tesla model S has an auto-pilot mode, which sends information about the places the car travels to, creating an ever growing map with data about the type of road, traffic conditions and other pertinent information. This is going to be true for any autonomous car if it has to become a viable means of transportation. This large-scale collection of data as promising as it is in terms of improving the way autonomous cars are able to understand their surroundings and adapt to them, it raises concerns about the privacy of the consumer. The information thus collected will inevitably be about the consumer’s personal details. This is problematic as the company can use the information for purposes not immediately considered to be for the benefit of the end user for instance, advertising, etc. The government has to make strict provisions making consumer’s informed consent mandatory for any way the data is to be used by the company. This data again has to be well-protected from cyber-attacks and hacks. The manufacturers should be required to maintain robust systems to ensure the safety of the information collected and conduct periodic tests to assess the workings of the system.

When accidents happen in normal circumstances, the driver is in most cases considered liable, but with the advent of autonomous cars fixing liability would become difficult. To be able to identify the events immediately preceding the accidents, German lawmakers have made it mandatory to have black boxes in all autonomous cars, a similar step can be taken here too. The black boxes will have information about when the driver took control of the car (if he did so) and what malfunction could have lead to the mishap. The liability would be placed on the manufacturer unless the consumer has not added any new feature through his own volition.

The possibilities which this technology presents in terms of positive outcomes is that it can lead to in say for instance in reducing the number of accidents, providing an accessible means of transportation for the old, disabled, etc. But, to harness the full potential of these vehicles we need to have a sound system of law which protects all stakeholders from the challenges which the introduction of these cars could pose.