Anatomy Of An Autonomous Vehicle Service Ecosystem
October 17, 2019 | In The Press
If you have seen one of the many schematic charts full of logos illustrating the autonomous vehicle ecosystem, you would be forgiven for being confused.
Most, like the one linked to in the above paragraph, dive deep into the layers of technology involved in enabling cars to drive themselves. It provides a nice summary for people in the industry (with good eyesight). To the layperson, however, this can add to the confusion about how autonomous vehicles work.
Also, it is important to note that the majority of the companies and the technologies represented only have to do with the vehicles. Many expect autonomous vehicles to help reduce urban congestion and pollution by moving more people with fewer cars. For this to happen, these vehicles will need to work together as efficient, coordinated fleets that deliver services. Autonomous services have an ecosystem of their own that can be broken down into six general categories.
It starts, of course, with a vehicle, which is the first layer of the autonomous service ecosystem. While the myriad of technologies inside the vehicle enables the vehicle to drive itself, it is fundamentally no different than any other vehicle. The vehicle is a chassis with wheels and seats and is designed to carry passengers.
The next layer in the ecosystem is the self-driving technology that drives the vehicle. This technology — a computer that receives input from sensors, cameras, radars, maps, and other sources and makes decisions to accelerate, steer, turn, and brake — takes the place of a human driver. Some vehicle manufacturers like Tesla are also creating their own tech stacks. Other tech companies like Waymo are developing the technology to be fitted to third-party vehicles.
As soon as you have more than one vehicle, you have a fleet. Most fleets of any size are monitored and managed using asset management software. This software tracks vehicle metrics like miles driven, speeds, fuel levels, and battery and brake levels. It is used to oversee vehicle and driver performance. These businesses have been around for decades and are transitioning to support autonomous vehicles as well as conventional fleets.
Orchestration is what turns a fleet into a service (it’s also the particular aspect of the autonomous vehicle service ecosystem that my company deals with). Fleet orchestration is different from asset management in that this technology is where the services that the vehicles provide are configured and managed. Fleet orchestration software serves several functions including:
* Service configuration: The software needs to know the number of vehicles, where they are located, how many passengers they can hold, etc.
* Service level definition: Fleet operators need to define maximum wait times and ride times for passengers and the criteria to be used when determining which vehicle to send on a mission.
* Operations automation: Autonomous vehicles need to be told where to go, when to go, and how to get there as efficiently as possible. This is automated dispatching, rider-vehicle matching and routing. The vehicles also need to be told where to go when not in service. This requires demand management — aggregating historical and real-time demand to continually position and reposition vehicles to maximize utilization. Empty vehicles lose money, so this kind of optimization is critical.
Mobility service providers are the businesses that operate the services that the autonomous vehicles provide. These businesses own or lease vehicles, accept bookings and payment, and use fleet management and fleet orchestration technologies to run the services. These could be:
* Conventional public transport operators that replace traditional buses with autonomous ones.
* Ride-hailing companies that transition from human drivers to so-called robotaxis.
* Private businesses or campuses that use autonomous shuttles to move people around their facilities.
MaaS platforms are mobile apps that aggregate multiple mobility services to provide a one-stop, door-to-door trip planning and booking service. For example, if a traveler enters a destination, a MaaS app would display a variety of options — a robotaxi, scooter, bike, public bus or train, etc. The traveler would be able to select the options he or she wants, book the entire trip with a click and be guided from door to door. To date in the United States, there are no full-fledged MaaS platforms. The best example is Europe’s MaaS Alliance, a public-private partnership that offers the Whim app. Whim allows subscribers to plan and book trips using just about any available mode of transit, with options sorted by cost and convenience.
Of course, fully autonomous services have yet to be delivered. Businesses are angling to capture as much of the value chain as possible. For example, if a vehicle manufacturer can develop its own self-driving tech, manages the assets, and orchestrates and delivers the service, it will retain the lion’s share of the revenue. Most expect MaaS platforms to be run by neutral third parties, as they will have to give travelers equal access to all available services. Uber and Lyft have taken steps to add public transport schedules and bookings to their apps in some cities. It remains to be seen if they can add more providers — essentially competitors — and provide a level playing field for all.
The autonomous service ecosystem continues to evolve, and how full-blown services will be delivered remains to be seen. But it is important to remember that vehicles alone do not provide services and that there is a different ecosystem involved in delivering autonomous services than the one that’s involved in building vehicles themselves.