Driverless Cars: Enabling the Transition of the ‘Driving Society’ to a ‘Driverless Society’ – Part II
Continuing our conversation with Mahbubul Alam, CTO/CMO of Movimento, an Aptiv company, we bring you the second part of our Driverless Car series, with interesting elements on data & analytics and the future of the driverless society.
Q. What is the role of data & analytics in improving performance for driverless cars?
Ans. Data & analytics have become an integral part of performance improvement for these cars.
Data is of various types and each type holds a different value. Analytics, in turn, can be primarily categorized into two types – (1) Real time analytics and (2) Non-real time analytics.
- Real time analytics: Data often loses its value with time, sometimes almost instantaneously. Hence the need for data to be captured in real-time and updated for effective utilization. There are two types of data that are relevant in real-time – one that enables vehicle safety for automated driving and the other that enables the creation of a personalized automotive experience. When it comes to safety, decisions have to be made in nanoseconds. For example, the computer vision of automated cars, that uses LIDAR and RADAR for creating images of the objects on the road, is being used along with deep learning algorithms to make instantaneous decisions about vehicle control and maneuvering. Did you know that $73 billion is wasted in the US annually just to find a parking spot? With real-time data, drivers can be provided with information about vacant parking spots based on their exact location. This would not only save time and reduce stress but it would also add value to businesses.
- Non-real time analytics: It focuses on data that is recorded on a regular basis and provides insights into driver behavior, wear-and-tear, diagnostics and the overall functioning of the vehicle. This data can be further used to proactively schedule vehicle maintenance or cyber security related software updates. Non-real time analytics can also provide updates to users to help them analyze their driving, their carbon footprint and so on.
Q. What about cybersecurity concerns?
Ans. One of the biggest challenges in driverless cars is cybersecurity. Cyber attacks, including hijacking, ransomware attacks, terrorism and data theft, have significantly increased over the recent years. The best way to prevent cyber attacks is to have a national threat defense system for the automotive industry. We have a defense mechanism for operations on land, air and water. The Internet should be added as the fourth dimension of national security. The automotive companies should come together and share data to collaborate on cybersecurity. Auto OEMs can have a shared database to maintain a collective defense system.
Q. How can we protect our personal travel data with driverless cars?
Ans. Driverless cars are associated with two types of data. The first is the vehicle-generated data that includes data on vehicle health, machine wear and tear, etc. The second is PII – Personal Identifiable Information – which includes data that can be traced back to the user. In the case of renting or sharing a car, the ability to erase all of your personal data must be present. The data profiles can be stored in the cloud and provided on-demand. If you own the car, better in-vehicle security is required so that you can safeguard your data profile with biometric identification.
OEMs should proactively mark the data as personal data or machine data. Often, the end users do not even know that their data is being captured. This should be avoided. OEMs should ask the end users for their explicit consent. The end users should be told how their data will be used and monetized. We need infrastructural technology that will allow government inspections to regulate the usage of data. One of these underlining technologies could be Blockchain, where there are distributed ledgers/database that establish a transparent data-sharing system for how the data is being used by the OEMs, thereby making them accountable to the end users.
Q. How soon will driverless cars be available in the market? Any comment on that?
Ans. In the last blog, we talked about the different levels of automated driving. Technically, we have reached quite close to the Level 5 automation where there is no need for a steering wheel, just that the driver has to be alert in order to take over the vehicle in emergency situations.
In most probability, the driverless car is going to come to vertical markets first. They will see a beginning in buses and large vehicles. They will be seen in vertical markets like trucking long haul on freeways, airports, university campus areas and between different hospital buildings to accelerate the mobility of doctors and medical staff. The automated drive can be much more easily implemented and accepted in these areas as the vehicles will drive below 30 miles/hour, which allows for better reaction time. Also, these areas are not as densely populated or complicated as urban traffic zones.
It will slowly extend to the urban areas but the automated drive will be restricted with speed in city centers and traffic zones. Truly automated drive for short distances at a lower speed will enable humans to trust the machine. The true driverless world with no steering wheel will be post-2030, where there will be new legislation and a new generation of millennials who will be much more confident in driverless vehicles.
Q. How do you envision the transition from today’s driving society to the future’s driverless society?
Ans. The true challenge for driverless cars is in dense urban areas where traffic rules are seldom followed and pedestrians are seen jaywalking. Instant human decisions are required in these situations to avoid any accidents. Autonomous driving on the freeway is already being implemented and will become a popular feature in all cars in the near future. It is easy to have automated drive on the freeway because the roads and the lane marks are clear, but, for example, if there is snow on the freeway, the driverless feature of the vehicle will be deactivated and the human has to take over the driving.
There will be a mixed world of fully autonomous cars and partial autonomous cars for a long time because of vehicle lifecycle, investment, certain road conditions and the location itself.
The adoption of driverless cars is going to increase from the two extreme ends of the demographics. In the last blog, we talked about giving elderly people the mobility to lead an independent life with driverless cars. On the other hand, younger generations will be more interested in on-demand driverless vehicles rather than buying them. These two demographics will be the early adopters. The rest of the population will start getting comfortable with these cars by using them at airports, hospitals and university/office campuses. Even though initially there will be some failures, they will be overcome with rapid fixes.
While all of us await the convenience of a driverless car, we should, in the meantime, work on making our roads better and safer, for when the automated drive technology is ready for adoption, we can be ready to accept and implement it without a glitch.
Our next post for the Driverless Cars series will focus on how Movimento, in collaboration with ValueLabs is bringing technological advances to the automotive industry.