A regular feature of the Fleet 500 issue of Automotive Fleet is a comprehensive listing of the Top 300 fleets in the U.S., with each fleet’s composition broken out by vehicle segment, such as the number of cars, vans, crossovers, and so on.
One category that will be added in the future is the number of autonomous vehicles (AVs) in each fleet. While that day is coming, I believe volume penetration of fleets by autonomous vehicles will take much longer to occur than what is predicted in today’s optimistic forecasts. Conceptually, autonomous vehicles are technologically feasible, but, as they say, the devil is in the details. One thing is certain, as we trail blaze new ground, so too will we trail blaze new problems. None of these obstacles will be insurmountable, but they will act as ballast slowing the speed of deployment. Below are a few “obstacles” that need to be addressed:
Recognition of Hand Gestures: Autonomous vehicles find it difficult to decipher the meaning of hand gestures by a driver or pedestrian giving it the right of way. Conversely, an AV runs the risk of misinterpreting a causal hand motion and erroneously determine it has been given the right of way. Understanding gestures is a major safety issue since it is used in a construction zone, by police officers directing traffic in an intersection with a disabled stop light, officers redirecting traffic around a vehicle accident, and at major public venues, such as concerts and sporting events.
Difficulty in Driving in Snow: AVs are programed to recognize various road landmarks, such as lane dividers, turning only lanes, or when two lanes merge into a single lane. But when covered by snow, these road markings are obscured or hidden. Similarly, when snow or road grime cover vehicle sensors, it “blinds” an AV to its surroundings impeding autonomous navigation.
Driving through Standing Water or Flooded Street: Similar to the issues encountered when driving in snow, standing water likewise obscures road markings impeding AV travel.
Inability of AVs to Effectively Communicate with Human-Controlled Vehicles: There will be a long transition period when AVs share the road with human-controlled vehicles. This will create a period of incongruity. For instance, a vehicle horn is used as a communication device providing a cautionary warning. Can an autonomous vehicle hear a horn? What about the colloquial rules of the road? For instance, will an AV know what is meant when a vehicle in the oncoming lane flashes its lights to yield to it the right of way? Flashing lights are also ambiguous because it can also signify displeasure. Can an AV distinguish between the two? What about ambulances and fire engines that in an emergency do not follow the rules of the road, such as driving through red lights? Ultimately, vehicle-to-vehicle technology will be embedded in emergency vehicles, but a fire truck can be in service for as long as 20 or more years before it is replaced.
Decades needed to Develop Case: AVs will necessitate the creation of an entirely new body of case law that will take decades to adjudicate legal precedents. It is generally accepted that if an AV is forced to make a choice between hurting a single person versus many individuals, the AV will be programmed to take the course of least damage. However, could a lawyer representing the aggrieved family argue premediated manslaughter (or even second degree murder) because the AV was programmed to intentionally hit a person? Unfortunately, the death of a pedestrian by an Uber AV won’t be the last person killed by an AV. What will be