Going beyond transformation: An autonomous and digital future
Oct 6, 2021 by Dr. Ulrich Wurstbauer
Currently, automotive OEM assembly plants are in a major transition process, shifting to the production of electrified vehicles. This moment calls for large scale investments and presents a great chance to modernize other big areas of the plants, too: Enabling the autonomous driving of vehicles from production line to other parts of the plant premises for example. But what are the requirements and gained benefits of such a transition?
Picturing an OEM assembly plant brings – for most people – images of autonomous robots bending and welding parts together, with just a few highly skilled people performing manual tasks in fancy equipment, like exoskeletons. While this image is not far from the truth, once a car is fully assembled and reaches the end of the assembly line, the situation completely turns on its head. When complete, every car requires an action which nearly everyone knows very well – a human steps in, starts the engine and drives the car to its parking position. From there, it is picked up and loaded on to trucks, trains or ships for delivery. Due to the high cycle rates of the plants (~2cars/minute), this final, human-powered, process often requires more than 50 people per plant. Emptying the end of the production conveyer then, is a substantial cost item in the operations of a car plant.
This huge human labor effort is actually a lot more complicated than it may first seem: Firstly, because modern plants are highly flexible on the type of vehicles they can produce (and they operate 24/7). All different types of vehicles are produced on the same line one after another – in an extreme scenario, it might occur that a combustion engine car is produced right after an electrified one, followed by a fully-fledged autonomous SAE level 3 car. The human driver responsible for moving such vehicles from the end of the line, is experienced enough to simply get into any of these cars and drives it away. Secondly, driving the vehicle to the parking spot is never the same twice – different moving parts, people, other vehicles and obstacles may occur between the production line and the parking spot on every journey.
The cost and risk of this human-reliant final step can be alleviated through automation, and initiating this change coincides perfectly with the current modernization of OEM plants. It’s also no mean feat though, as not every car will have all the required sensor and compute capabilities on board to drive autonomously, and an autonomous system for the whole line would also have to be able to cover all the different scenarios arising from other vehicles and pedestrians etc. on the plant grounds.
Infrastructure can bridge the gap(s) of autonomously driving and parking the manufactured vehicles at the plant grounds. Our solution relies on automating driving tasks generation through the backend. Simply speaking, the solution utilizes sensorics placed in the surrounding infrastructure to sense the environment, localize the vehicle and compute the required driving commands with autonomous driving algorithms on the backend. Utilizing the OEM’s connected vehicle platforms, driving commands are transmitted to the vehicles – all the vehicles have to do, is execute the driving commands. This system allows for more efficient movement and parking of vehicles, with no driver delay and reduced space between parking spaces.
This solution overcomes the aforementioned production line challenges, as all required sensorics are placed outside the vehicles1. Another huge advantage, is that the whole autonomous driving (AD) computation is outside the vehicles, so non-AD cars can be driven as well (a very important factor for many OEMs, considering the percentage of fully equipped AD vehicles on production lines2).
In addition to nullifying the challenge of moving non-AD vehicles, our solution also provides the advantage of a bird's eye view of the whole plant’s outbound logistics. This digitalized view extends the plant management’s possibilities from knowing where vehicles are stored, to actively and automatically orchestrating vehicle movements.
The momentum of switching to electrification can, in parallel, bring the opportunity to address other issues too. Our solution can supplant missing in-vehicle automation capabilities to close the missing autonomy at the end of production lines, bringing substantial savings to the benefit of plant owners. It’s an automatic advantage.
1 Of course in vehicle sensorics can optionally added as well for enriching the available data landscape. 2 It is worth to think about this also from the final cost point of such cars to the end client.