Autonomous Vehicles: A Horse of a Different Color (Part 1 of 4)
The driverless car is coming. In this series, we'll look at the implications to data collection and privacy, to data analytics, and to business and society in general.
- By Barry Devlin
- July 11, 2016
"Why this time is different: going from horses to cars was fine for people, but now we're in the position of the horses" Tweet by Ben Kinnard (@CaptainKinnard), 6 June 2016
The first half of the 20th century saw the rapid and complete displacement of horses by cars, trucks, and tractors on the streets and farms of the western world. Kinnard's pithy tweet compares their fate to our own in the transition now beginning from automobiles to autonomous vehicles: the changing roles of human intelligence and machine intelligence as we humans are displaced from the driving seat and disconnected from the reality of travel.
In this series I'll examine the growth of data collection, analytics, and artificial intelligence in driverless cars and its implications for people, privacy, business, and society in general.
Dials, Dashboards, and Data
By introducing automobiles early in the 20th century, we did more than simply replace horse power with horsepower. We broke the bond between human and horse, and the person driving assumed full responsibility for control of the vehicle. Information, including data about the internal operation of the vehicle, became central to the responsiveness demanded when driving a car.
It's exactly 20 years since all new cars in the U.S. were required to be fitted with a set of basic sensors and onboard diagnostics based on the OBD-II standard. Before this requirement, automobiles had already begun to gather data, albeit of the most basic kind, such as engine temperature and oil pressure. As sensors became more pervasive and onboard controllers more sophisticated, automobiles gained the first glimmer of sentience: they could monitor traction on the road to prevent skidding or wheel-locking and they could shut themselves down before they could be damaged by overheating.
Over the past few years, the information gathered and used by vehicles -- location and speed, turning and braking forces, fuel consumption, external weather and traffic conditions, local speed limits, use of seat belts -- has grown in variety, velocity, and volume (in a real example of much-hyped marketing speak)). With this combination of internally and externally sourced data, faster processors, and improving algorithms, the focus has shifted from engine monitoring and preventive vehicle maintenance to inferring driver behavior and supporting driver performance.
Autonomous vehicles demand further information: LiDAR maps and video cameras observe the surroundings, map updates and real-time routing are downloaded from the cloud. Vehicles start to "see" and "feel" the environment. The automobile finally begins to match the horse's awareness of its surroundings.
A Sense of Separation
However, as vehicle intelligence has grown a feeling of separation between driver and vehicle has emerged, quite unlike the old relationship between human and horse. We now depend on the car to interpret its electromechanical state and tell us what to do.
My father could simply listen to the sound of his beloved engine to decide what maintenance was needed. I can only read the lights and gauges, although I still get enough information to feel a connection and a sense of ownership. As operational data and external information is further internalized to the vehicle, my children are disassociated from the car. They care only about getting from A to B; how becomes irrelevant as the satellite navigation system purrs its directions or the Uber driver obliges. As the separation grows, the desire for car ownership diminishes.
Thus, the scene is set for the arrival of the autonomous vehicle. The data and information needed is widely available and the necessary sensors increasingly affordable. The cloud is pervasive and connectivity always-on to provide any new information required. The processing power is more than adequate for real-time analysis and control. Storage is cheap. Artificial intelligence and deep learning are advancing toward dependable decision making in ever-changing and unpredictable driving situations. Mobile apps are already prevalent for ride-sharing and car rental. Government and industry are engaged; trials proceed apace.
The availability of new data in novel combinations always induces changes in society and business. Opportunities arise for new businesses and business processes, while old models and approaches are threatened. In Part 2 of this series I'll explore the new data autonomous vehicles can collect and who will use it.
Dr. Barry Devlin is among the foremost authorities on business insight and one of the founders of data warehousing in 1988. With over 40 years of IT experience, including 20 years with IBM as a Distinguished Engineer, he is a widely respected analyst, consultant, lecturer, and author of “Data Warehouse -- from Architecture to Implementation" and "Business unIntelligence--Insight and Innovation beyond Analytics and Big Data" as well as numerous white papers. As founder and principal of 9sight Consulting, Devlin develops new architectural models and provides international, strategic thought leadership from Cornwall. His latest book, "Cloud Data Warehousing, Volume I: Architecting Data Warehouse, Lakehouse, Mesh, and Fabric," is now available.