On March 15, 2017, Mr. Marco Cava wrote for the USA Today about the cloudy future of autonomous cars. Capturing the concerns, Mr. Cava wrote, “While there remains a general feeling of technological inevitability about self-driving vehicles, industry experts remain concerned about the shift’s looming logistical, legal and social ramifications.” Indeed, there is an inevitability to autonomy in vehicles, but the future requires a lot of collaboration.
Autonomous vehicles are already on the road and functioning. There has been tremendous value to the developers gaining experience by pushing their vehicles out into the real world. Controlled laboratory settings only go so far before testing has to move to the operational environment. This absolutely must be done as safely as possible, but there are simply too many variables to test behind a fence. The laboratory must be extended. However, the unveiling of autonomous cars in urban population centers are high-stakes gambles in publicity vs. risk. Rolling out the future in rural Iowa doesn’t resonate loudly for Silicon Valley and venture capitalists seeking to establish a strong, early presence in the autonomous vehicle field. Driverless cars in San Francisco gain publicity and provide a platform for promotion. Driverless cars in Iowa gain experience and improvements, but make bringing in more money difficult. It’s a tradeoff of making money and making the technology work.
Big automakers have been taking the slower approach. Their incremental steps have included such technologies as braking assists, adaptive cruise controls, and automated parking. This isn’t to say they don’t have the ability to make a fully autonomous vehicle. Mr. Cava quotes Ford’s Bill Ford saying, “The [autonomous] vehicles will be ready by 2021.” We already know the cars can be built. But, it’s hard to sell vehicles with all of the appendages created by optical and radar systems we see in current fully-autonomous vehicles. Sleek and stylish depends heavily on shrinking and fusing the inputs required to create a computer-generated model of the world around the car. Small size coupled with technological power is the bane of affordability. Ford and other major automakers aren’t looking for a niche market, they want volume. Mercedes can move cars costing in the six-figures, but this isn’t Ford’s market place.
To gain market penetration in volume, some very large hurdles have to be overcome. A key limiter is the infrastructure required for the vehicle to ascertain where it is in both time (velocity/acceleration) and space (relative location) and make independent decisions about how these vectors are changing in relation to fixed and moving objects. In the lab, this is an academic exercise, but on Main Street those decisions must be context-based. Again, quoting Mr. Ford, we are forced to consider “the ethics of autonomy” and how the vehicle chooses between two evils in an emergency. Even ignoring for now the vehicle’s artificial intelligence having to decide between a hitting a car or hitting a school bus, approaching context-based decision-making in a vehicle requires an incredible amount of both stored and updating information that is being parsed continuously. These information sources must exist passively and actively outside the vehicle (as well as in) and those outside sources can require significant funding, even if it’s just painting new lines on the road.
The infrastructure issue makes autonomous cars a national issue rather than a local one. Common standards between states are required and the costs of meeting those stands will vary. California would potentially face and exponentially higher bill in applying infrastructure standards than say Wyoming. Not to mention that a low-population, ranching-based state like Wyoming may not be willing to spend its tax base on smart-car infrastructure. Federal dollars are critical. The current spotlight on the U.S. transportation infrastructure has been negative and politicians are feeling pressure to alleviate the strains on the aged roads and bridges. A collaborative effort between the autonomous vehicle developers and the planners for national-level refurbishment as early as possible would potentially alleviate much after-the-fact upgrades for an automated future.
The technology is exciting and brings out the hopefulness for a better future in transportation. Overcoming the infrastructure issue is one of the most difficult challenges facing mass acceptance of autonomous vehicles. Finding a technological way around it may actually be easier than getting politicians to agree on a standard and then paying for it.
Cava, M. D. (2017, March 29). Sunny autonomous car futures faces some clouds. Retrieved April 03, 2017, from https://www.usatoday.com/story/tech/news/2017/03/15/sbsw-austin-autonomous-cars-self-driving-future-roadblocks/99196152/