Fully self-propelled passenger cars are not "just around the corner". While well-capitalized leaders funded by corporations, multimillion-dollar venture capital funds or advertising revenues. They are in a more stable financial terrain, many other autonomous vehicle start-up companies may be looking for the exit ramp.
Without a clear path towards funds outside venture capital. Start-up companies face two options: 1) being acquired by talent and technology or 2) closing the store.
Cruise and Argo IA were great initial starts. Daimler Trucks acquired Torc Robotics (which did not follow the VC-startup model). And nuTonomy was marketed as a $ 450 million acquisition by Delphi / Aptiv.
But the most recent valuations backed by VC for some AV companies have stagnated at or below the $ 450 million mark. What does not give much advantage over their previous valuations at the height of the fervor AV. Without further advantages, it is more likely that many touring car companies close the store.
The start-up of full-stack autonomous passenger vehicles is dead.
Autonomous vehicles for other strata
The autonomy projects of passenger cars attracted a lot of capital and talent in the last decade and produced enormous technological advances in the autonomous perception, planning and control of the route. What happens with talent and technology when the passenger's AV bubble bursts?
Well, there are more vehicles than just passenger cars. The DARPA Grand Challenge held more than a decade ago is cited as the catalyst behind the GoogleX self-driving project and the explosion of audio-visual systems for passenger cars.
Advances made during the challenges also extended to off-road vehicles. Since then, autonomous vehicles have been developed and deployed both in defense and commercially in agriculture and large-scale mining.
It is widely observed that industrial, agricultural, construction and mining applications are more suitable for short-term autonomy. There are defined automation tasks with a clear return on investment. Fewer man-machine interactions and geo-fencing areas that limit operational and safety requirements. It is simply more controlled environments than on the streets of cities.
Automation is the key to autonomous vehicles
Automation can also help compensate for the shortage of critical labor. It is difficult to attract labor to remote mines in the middle of large deserts. The shortage of labor for agriculture adds enormous uncertainty to farmers who do not know if they can prepare and harvest their crops in short periods of time.
With the help of DARPA participants, Caterpillar developed semi-autonomous and fully autonomous transport trucks and announced that it has transported more than 1 billion tonnes of material. A day later, Komatsu announced that it had reached the milestone of 2 billion tons. These transport trucks are the size of a house.
John Deere, Case IH, New Holland and others have developed semi and fully autonomous tractors on their own and with the help of R & D companies. Most of these programs have existed for more than a decade. But the pace of technological progress is very similar to that of recent start-up efforts.
What is coming?
From our point of view as investors. We believe that we will see a similar impact of the AV bubble of passenger cars in the industrial, agricultural, construction and mining sectors.
This will improve the existing autonomous programs, open new cases of ROI use in those sectors and remodel the business model of autonomous vehicles in some of the sectors as smaller players gain access to the best talents and technology.
The most significant technologies that will be extended to the off-road vehicle market are the perception of the machine. Reinforcement learning for more complex robotic motion planning and functionally safe and mission-critical engineering requirements.
Perception systems used in mining and agricultural vehicles are not as expensive as passenger cars. The price tags of some of the CAT 700 series trucks exceed $ 5,000,000.
These vehicles are equipped with a robust lidar, radar, cameras, etc. Mainly to raise awareness about safety. The costs of these systems will decrease thanks to the low cost designs of the sensors driven by the automotive market.
Camera-based inference will allow these vehicles to better understand the elements of their environment. This will allow them to perform more complex tasks and navigation operations.
The fusion of sensors can allow agricultural vehicles to use optimal inputs in the fields. Or in mining vehicles to understand the characteristics of the ore and increase the productivity per scoop.
The learning by the cars is the key
Reinforcement learning allows operators to "teach" algorithms to perform complex tasks and create new use cases that require complex robotic performance.
These use cases could be the harvest of more than one acre of land, the movement of land in the place, the collection and placement of construction equipment for the start-up and much more. These robotic applications can be integrated over existing autonomous mobility platforms.
The most important criterion for these start-ups is an inflexible approach to robustness and safety. Autonomy only reaches its maximum potential if the solution works with minimal downtime and improves security (which is also linked to the costs of equipment replacement, workers' compensation and insurance).
Recognizing these trends, we have made an investment in an AV start-up that is deploying stand-alone systems in Bobcat skid steers and excavators in construction. Working with large mining operations to automate all vehicles at the mine site.
We have also invested in an early stage agricultural robotics company that automates applications in the field that have not been affected by automation so far.
This is just the beginning. There are many more opportunities in off-road autonomy, and we continue our search for companies in other off-road applications.