Autonomous driving is an eventuality that the automotive industry is looking to reach, and while fully-autonomous vehicles for regular transit may be far away, technologies to enable it have been in development in recent years
The degrees of automated driving largely adopted universally are those prescribed by the Society for Automotive Engineers International (SAE International) under the SAE J3016 mandate. It defines vehicle autonomy from Level 0 to Level 5, in order of the intensity of self-driving technologies incorporated into the vehicle.
From a broader perspective, Levels 0-2 include the need for the human in the vehicle to drive and constantly supervise the various parameters of the vehicle, environment and all-round conditions. The support features offered in these forms of autonomous driving can range from limited warnings alone, right up to steering and braking/ acceleration assistance. Some of the features offered in Level 0-2 are automatic emergency braking, blind spot warning, lane departure warning, lane centring and adaptive cruise control.
However, the real essence of automated driving kicks in during SAE Levels 3-5, wherein the human in the vehicle does not carry out the task of driving, once the automated features are engaged. Level 3 will still require the driver to take over, when the automated driving features requests for the same, whereas the features in Levels 4 and 5 will not require the driver to take over the vehicle at any given point of time. Some of the features of automation found in Level 3-5 vehicles include traffic jam chauffeur, local driverless taxi, absence of steering and pedals, in that order. The features in Level 3 and Level 4 automated driving can take control of the vehicle in limited conditions and not operate unless all the prescribed conditions are met. Meanwhile, Level 5 is the ultimate form of automated driving, wherein the vehicle is navigating itself autonomously across all possible conditions.
The enablement of autonomous driving depends strongly on the technologies that back it up and make it a system that can be brought onto the roads in the future. It is made up of technologies such as advanced driver assistance systems (ADAS), connected ecosystem and sensor fusion. ADAS serves as the main foundation to a strong and safer automated driving system, which presents itself across all levels of autonomy. ADAS features such as automated emergency braking, adaptive cruise control, lane departure warning, blind-spot detection, forward collision warning, traffic sign recognition, tyre pressure monitoring system, night vision, pedestrian detection and parking assistance enable automated driving.
These ADAS features are infused into vehicles with different levels of sophistication to enable the desired level of autonomy in driving. These systems collect data regarding the vehicle, driver and surroundings through various receivers such as camera, radar, LIDAR and navigation maps to then analyse the situation and take appropriate decisions, in terms of driving. This shows that the vehicle needs to be connected to other vehicles on the road, as well as to the overall infrastructure for different types of information related to the road and conditions. The degree to which the vehicle and its systems are connected to each another, as well as to other vehicles and multiple information-providers is critical. Such information needs to be analysed, and the resulting steps would facilitate the decision-making aspect of autonomous vehicles.
When we look within the vehicle, the need for the integration of multiple systems of vision, radar, LIDAR and maps, among others, becomes apparent. This essentially means that three main sensors (camera, radar, LIDAR) and related control units need to communicate with each other to grasp the environment as well as react to it by taking relevant actions that would ensure safer driving.
This brings in the requirement for sensor fusion. This technology that enables collection of information with sensors, including actual path, traffic jams and obstacles, will be shared between IoT connected cars for vehicle-to-vehicle communication, which improves driving automation. The main reason for sensor fusion is that all these sensors have their own limitations under different conditions, and their unified solutions would drive in a higher level of safety to the autonomous vehicle sphere.
The adoption of ADAS systems in vehicles is definitely on the rise, but the practical usage of these systems in the real-world calls for development of overall road infrastructure as well as that of vehicle connectivity. In addition, the mandates of regulation with regards to safety features in vehicles would also help enhance the application and adoption of such systems that help bring in more automated driving functions onto the roads. The growth for technologies of vehicle-to-everything (V2X) connectivity and sensor fusion will back the ADAS solutions needed for the different levels of autonomous driving. In addition, the high costs involved in these sensor technologies, especially that of LIDAR systems could act as deterrents in the deployment of automated transportation solutions across markets.
Therefore, it could be said that autonomous driving Levels 1 and 2, and to some extent even Level 3, are likely to be deployed slowly in automotive markets across the globe over the next few years. For starters, the deployment of these initial levels of automated driving may be seen in the mid to high-level personal passenger vehicle space. This is because these segments of the automotive industry are open to adopting new technologies at an additional cost, much before the mass segment does. These segments generally pay a premium for technologies that they feel bring in additional value to the product, in terms of safety and technology.
The adoption of Level 4 autonomy and eventually Level 5 cannot be expected any time soon in the future. There may be stray cases of fully-autonomous vehicles plying on open roads for personal use in pockets of small countries across Europe, US and Asia. However, the case of such technologies being put to use in restricted and repetitive automotive roles is much higher. For example, the use of autonomous trucks in mines, which have fixed routes and repetitive tasks, or that of autonomous tractors to plough a set area of enclosed land will be seen first.
The use of autonomous vehicles on public roads alongside human-driven vehicles may not be seen on a large scale for a couple of decades at least. Even when this situation takes place, it could be in the roll-out of autonomous buses or taxis that could be restricted to specific lanes or cordoned off areas of the highway, in order to avoid any conflict with regular vehicles that are not autonomous in nature. The high vehicle population themselves could also act as a negative factor for autonomous vehicles, and therefore increased levels of shared mobility could, in fact, address a major challenge for automated driving solutions.
Fully-automated driving will one day become the norm. However, the journey towards this goal will still be a long one, with the path being filled with innovations that will be showcased in bits and pieces. There is much development that has, is and needs to be carried out with regards to autonomous vehicles, since safety is of paramount importance.
The negative impacts of underdeveloped autonomous vehicles are severe, and could end in the loss of life and property, not to mention the steep levels of liability issues to be addressed thereafter. The various systems and technologies, and supporting infrastructure and connectivity for vehicle autonomy are witnessing increased levels of engineering and R&D, and this is what will bring this form of future mobility to the road.
(Inputs from BMW, Sasken Technologies, Synopsys, ITransition)
TEXT: Naveen Arul