1) The vehicle must successfully identify roadways from ‘non-roadways.’
This may be easier said than done… At what point is the gravel roadway present and not? Is there a size of pothole that interrupts the concrete or asphalt that will compel the vehicle to stop? How about snowy roadways? The adaptation demands are diverse.
2) The vehicle artificial intelligence must be able to navigate subjective obstacles.
If the majority of vehicles are still operated by an individual, then the artificial intelligence must be sufficiently capable of identifying the vehicle intermittently drifting over the lanes, because it’s driver is busy talking on their phone, to counter the non-orderly and erratic driving patterns of humans. Furthermore, let’s not forget the dangers of animal crossings, namely deer, that cross with little warning and are large enough to cause serious danger to the passenger occupants.
While the costs of human accidents may merit companies to develop the technology, it remains to be seen just how they will react when the first vehicular death results from the autonomously driven vehicle. All of these individual companies have much deeper pockets than most individuals AND how sizable will the damage awards be when the first pedestrian is run down by one of these vehicles? The passenger of the vehicle will be hardly responsible, they were not in control of its operation.
Here’s an alternative scenario… tankers and container ships…
The BBC reports that Rolls Royce has released a concept for autonomous or remotely driven ships. Considering the obstacles to vehicles, it appears more probable that ships are to operate on the high seas before individual passenger vehicles.
Let’s consider it in turn…
1) Waterway vs. Non-waterway… Automobiles and trucks operate on a very narrow range of roadways with little fluctuation in height (small pothole = fine, large/deep pothole = not fine, vehicle must stop or elude). The waters that a ships navigates are much more predictable. While the tides and weather may fluctuate depths and conditions, depth-finders and ocean charts provide the basis to think this is not an insurmountable challenge.
2) Artificial Intelligence (AI) and subjective obstacles… Here the challenge is more weather related. Ocean vessels tend to navigate familiar paths. The familiarity works in favor of developing artificial intelligence capacities capable of navigating the oceans and ports. While there are plenty of ships on the water that navigate to just a few destinations, the same cannot be said for the millions of car, truck, bus, motorcycle, and pedestrian obstacles that an AI driven vehicle must adapt to in order to avoid collision.
3) Liability… Losing a container ship is immensely expensive and a risk that insurance firms must incorporate into policy rates. However, this type of liability is nothing new. However, it may be offered that the orderly and systematic paths of container ships may increase the threat of piracy. This liability dimension is already a threat that insurance actuaries must incorporate, but if the ships operate more efficiently with few to no crew members then any increased costs from Jack Sparrow may be mitigated.
A ship operated autonomously, or remotely, does not necessarily entail zero crew members. Crew members may still ride with the ship to address mechanical failure or specifically for the role protection escort. The incentive may arise from the fact that the crew may be significantly reduced from current levels. The BBC article notes that, “According to Moore Stephens LLP, an industry consultant, crew costs account for 44% of total operating costs for a large container ship.” As a result, if the crew were reduced by even a few, then it is very likely to result in significant savings.
Given that there are few, if any, value added dimensions to the crew on a container ship, it follows that Rolls Royce’s concept deserves more interest and attention than the driver-less vehicle… However, a driver-less vehicle makes for a much sexier story than an autonomous container ship. How many of envision a day when we are chauffeured around over the prospect that a robot ship just deliver a shipping container of cell phones. I’ll still dream of the chauffeured vehicle… and my hydrogen powered car… and my jet pack, yet know that the more mundane is the more likely future reality.
Think I’m being overly skeptical… one need only look to some of the heads-up vision enhancement displays that GM has been tinkering with for years. They have yet to become standard equipment…