Managing Vehicle Availability to Improve Fleet Efficiency
January 19, 2021 | Company Blog
The Uber/Lyft phenomenon exposed the importance of fleet sizing and fleet efficiency in transportation services. The peer-to-peer ride-hailing business model flooded cities with vehicles to make on-demand rides as convenient as possible. Because peer-to-peer services don’t own vehicles and only pay drivers when a rider is on board, they could afford to oversupply markets. The result was dramatically increased congestion and pollution in cities and enormous financial loses for the drivers.
Since people are now used to convenient, on-demand services, new businesses are emerging to capitalize on the trend. But to avoid the downsides of the peer-to-peer model many are looking to own their fleets and make mobility services more efficient and profitable. This makes getting fleet sizing to meet demand essential.
Careful fleet planning and using the vehicles efficiently is critical to making sure the service carries as many paying passengers as possible without oversupplying (too many empty vehicles) or undersupplying (not enough to meet demand). Fleet efficiency is key to scaling a transportation business and delivering a sustainable service—especially when providers own their fleets and every mile/km, in particular empty miles, represents a cost.
Optimizing fleet efficiency means optimizing both fleet planning and vehicle dispatching/routing. The best dispatching/routing in the world cannot overcome a lack of or poor fleet planning.
Optimal fleet planning means having the right number of vehicles (fleet sizing), at the right time (shift planning) and at the right locations (vehicle positioning). These three factors combined are what mobility providers call vehicle “availability”—when and where a vehicle is available to perform rides. Getting vehicle availability right enables mobility providers to always have the right number of vehicles in the right place at the right time to meet demand.
But getting vehicle availability right is extremely complex. Demand intensity fluctuates throughout a given day and week, which means the fleet size will have to be flexible—hence the importance of shift planning. Demand locations also fluctuate which makes positioning an additional challenge. Having the right-sized fleet with vehicles too far from travelers can also result in unacceptable wait times and canceled bookings.
There are many factors that can impact vehicle shift planning. Issues to consider include “vehicle unavailability,” which is when vehicles are not usable due to maintenance, charging needs or planned cleaning. Prebooked rides also impact availability and have to be factored into planning. Prebooked rides play a role in both shift planning and vehicle positioning.
It might help to think of a mobility service as a restaurant chain. Demand fluctuates daily, weekly, and seasonally. Preparing too much food is wasteful and having too many waiters on hand when demand is low is a money-losing proposition. Having too few waiters when demand is high slows down service and frustrates customers. Also, demand profiles may differ at different restaurant locations and staff may have to work in different places throughout the week. Staff availability planning—having the right staff size, shift plans, and staff positioning is thus critical to success.
You can see how vehicle availability and fleet planning represents a significant challenge for fleet operators, especially for on-demand services. Optimizing fleet size, shifts, and positioning–continuously–is critical for mobility providers to offer predictable, efficient services. Accurate planning is the needed foundation on which dispatching and routing rely and thus provide the best possible service at the least possible cost.
As you can imagine, attempting to perform fleet planning for an on-demand mobility service directly with the physical fleet of vehicles is difficult as it can be very unstable and brittle. Planning with physical vehicles invokes too many uncertainties and is therefore hard to do for a long time horizon. Whenever something happens to a physical vehicle, the full plan has to be changed. These adjustments are extremely time consuming and may become expensive.
In the world of fixed-route services (buses, trains, etc.) virtual fleet planning is common. The routes are static and planners can model stop locations, frequencies, and demand profiles. Planners use a sequential, multi-level approach and do availability and shift planning with virtual resources first, on a tactical planning level, before assigning physical vehicles on an operational level, as late as possible before the start of the operation. Virtual fleet planning makes it possible to plan the resources’ availability while taking into account all of the last-minute and real-time constraints and variables.
Bestmile’s Fleet Orchestration Platform enables this kind of virtual fleet availability planning for on-demand services, using demand data and utilization requirements to determine the optimal fleet size, shift plans and vehicle positioning at all times. Pre-bookings are matched with virtual resources, which are automatically assigned to real vehicles before the start of the operation. If a vehicle becomes unavailable, the dispatcher or the system can trigger a reassignment of a new vehicle to the resource without having to redo the full plan. Thanks to this type of accurate multi-level and dynamic planning, the platform can automate dispatching, ride matching, and routing that delivers predictable ride times, wait times, and fleet efficiency.
The platform makes it possible to plan availability in both time and space–managing when vehicles will be available but also where they will be when they become available through the creation of “depots” where vehicles are parked when they start and end their shifts. These two locations can be different. At the end of a shift, the vehicles are directly routed to the destination that is optimal for the next shift.
As tools already exist to plan vehicle maintenance or the creation of vehicle availability might be the output of another process, Bestmile Platform has been designed “API-first”, with open interfaces that enable integration with external systems such as vehicle asset management as well as dedicated shift planning software. Bestmile APIs enable different ways to automate vehicle shift planning as tools and practices may differ by service type.
While dispatching is a fully automated process, and a lot of attention is on the dispatching and routing, managing fleet availability is often neglected. At Bestmile, we have therefore built a powerful multi-level fleet availability and vehicle shift planning technology with advanced automation and optimization capabilities to allow mobility providers move from manual planning to automated fleet availability planning deeply integrated with the fleet dispatching logic.