How to size passenger fleets using ‘What-if’ scenarios
Using simulation to make better fleet planning decisions
Passenger fleets operating in today’s challenging mobility environment need to maintain and improve operational efficiencies to remain competitive.
Importance of employing the right-sized fleet
Right-sizing fleets is commonly understood to mean having the correct number of vehicles to serve the demand for the service – but it involves far more than this. It is a plan to have suitable vehicles in the right place at the right time – a crucial step to ensuring fleet operations are cost-effective and sustainable.
Some fleet operators may realise that they spend too much money to have more vehicles than required, while others can improve their service with a more extensive fleet. A simulation-based right-sizing analysis enables us to understand current fleet performance better, and sets up fleets for future growth.
Set and measure KPIs
An essential first step in any performance assessment is understanding historical performance against the key indicators that drive your business.
Typical KPIs include passenger wait times, service fulfilment and time in service. If you have historical service data available, you’ll be able to set a baseline of how you are operating today, and compare it against alternate simulated futures.
Understand the demand for your service
A right-sized fleet is typically established for a specific demand. Demand refers to the trip patterns that the fleet vehicles are attempting to serve. They consist of trip origins (or pick up points), destinations (or drop off points) and times. Historical demand can be pulled from service data, whilst understanding demand for deployments in a new location is more challenging.
Using Immense, fleet operators can easily incorporate different demand data via CSV upload or random generation. Immense has also signed data partnerships with leading providers, enabling our users more direct access to real-world demand patterns.
Once operators have a baseline demand, users can begin to right-size the fleet to serve these trips.
Experiment with fleet configuration
With Immense, users build a fleet configuration by defining the key parameters – the number of vehicles, type of vehicles, allocation and routing strategies, recharge policies, and depot locations. Experiments can be built around these parameters. For example, you could run a set of fleet scenarios with one central depot, five distributed depots or ten micro-depots around London – or test the impact of vehicles going to recharge at 20% battery versus 15% or 10%. Once you have a set of scenarios, you can analyse the results to find the preferred fleet size and configuration based on what is important to you.
Users access Immense’s built-in analytics to view rapid insights into their experiments. For example, quickly see the number of vehicles required to provide a 5-minute maximum passenger wait time – or the impact of a new charging policy on the number of trips able to be served by a fleet across the day. Powerful simulation tools like Immense enable you to assess a wide array of scenarios and use the right-sizing analysis to prepare a blueprint for your operations. Exploring elements that aren’t linear significantly impact any fleet planning and strategy implementation, including a transition to electric vehicles.
Iterate and continue to improve
Right-sizing is part of continually improving operations, and operators need to be ready to adapt to changing circumstances. Operators need to evaluate and iterate their fleet capabilities regularly, especially as demand patterns continue to evolve. Our team suggests a minimum cadence of monthly evaluation alongside real-time tracking, and many fleet operators see value in weekly assessments to see what they could have achieved with a different fleet configuration.
Scenario analysis helps fleet operators to stress test their assumptions on trip patterns and fleet efficiency, aiding the development of a blueprint for sustainable and cost-effective operations going forward.