Scenario Comparison

Want to know what happens if you take two vehicles off the fleet, move a depot, or bid for a new contract? Optrak runs the alternative through the same planning engine you use every day and shows the two side by side - cost, CO₂, vehicle utilisation, driver hours, and how many plans would not have completed without spot-hire.

Why a Spreadsheet Answer Misleads

A strategic decision in fleet operations - fleet sizing, a depot move, a contract bid, a change of shift pattern - usually reaches for a spreadsheet. Estimated mileage, vehicle running costs, driver hours against averages. The spreadsheet gives a number, the number looks defensible in the planning meeting, and the decision goes ahead.

The trouble is that the model the daily planner works with is far richer than any spreadsheet. Loading constraints, congestion-aware journey times, driver-hours rules, tractor-trailer compatibility, time windows - none of them abstract cleanly into a few cells. So decisions taken on the spreadsheet basis tend to surprise on first contact with operations. The depot move that was meant to save £80k a year adds £40k instead, because the new geography breaks two profitable rounds. The contract that looked profitable will not fit the fleet without spot-hire three days a week.

The cost of doing nothing is harder to see but just as real. Decisions get deferred because the strategic planner cannot get a credible model out of operations. Contracts go unbid because the cost to serve is too uncertain. Fleet right-sizing sits undone for years because nobody can defend the cut without operational evidence. Filshill took two vehicles out of their fleet, worth over £120k a year, but only once the routing system gave them the grounds to defend it.

What you want instead is a scenario that runs at full operational fidelity against the real data, so the comparison is grounded in the same model the operation runs against every day. That is what scenario comparison is.

How It Works

A scenario is a modification to the live planning configuration: vehicles added or removed, depots opened or closed, contracts added, time windows changed, the fleet’s fuel mix varied. The optimiser runs each scenario over a set of representative weeks sampled across the year - peak, normal, seasonal - using the same data model, the same constraints and the same heuristic catalogue as operational planning. The output is a side-by-side comparison on the metrics that matter to the decision: total cost, CO₂, vehicle utilisation, driver hours, missed delivery windows, and feasibility (how many plans completed without spot-hire or a relaxed constraint).

Two plans for the same 208 orders, totalled in one table. Opening every customer’s delivery window needs one journey fewer, and so one lorry fewer, since each lorry goes out once. It also saves 232 km, 74 litres of diesel and 9 hours’ driving, and the 21 hours spent waiting for windows to open go altogether. Time at customers’ sites is 129 hours in both. The table does not show orders delivered: 197 of the 208 with each customer’s own window, 198 with the windows opened.

A comparison can be set up and run as a modelling exercise, or run directly by the team that owns the decision in the self-serve strategic workbench, which also has an experimental AI agent window that takes the same steps from a plain-language request.

Technical Detail

Scenario Configuration

A scenario is the live planning configuration with one or more changes applied. The changes can be any combination of:

• Fleet changes - add or remove vehicles, change vehicle classes, change fuel types.
• Site changes - open or close depots, change opening hours, add charging infrastructure.
• Contract changes - add or drop customer contracts, change service-level requirements, change time windows.
• Schedule changes - shift-pattern changes, six-day working, double-shifting.
• External-condition changes - winter consumption profiles, peak-season congestion, fuel-price shifts.

Period Selection

A scenario can comprise a single dataset or multiple related datasets. Take fleet configuration as an example: If we wanted to replace 2 fleet vehicles with BEVs, we would take multiple operational plans and create copies with the revised fleet, replanning them. Our scenario would then have multiple days of datapoints in the comparison table to ensure that the metrics were not just one off. In the workbench this could be achieved by selecting the datasets with and without BEVs and using the built-in pivot table to separate and summarise the data.

Comparison Metrics

Standard metrics:

• Cost to serve - total operational cost across the period.
• CO₂ to serve - total emissions across the period, with mixed-fleet handling for diesel/EV/HVO/CNG/LNG.
• Feasibility - how many plans completed without spot-hire or a relaxed constraint, and how many needed overflow handling.
• Vehicle utilisation - average payload utilisation across the fleet.
• Driver hours - total used against available, including overtime.
• OTIF - modelled on-time-in-full performance against time windows.

Custom metrics, for a decision that turns on something specific (tonne-kilometres delivered per hour of overtime, cost per drop on a particular contract), are on the roadmap.

Worked Example: A Fleet-Reduction Decision

A wholesale distributor running 30 vehicles is asked by the FD whether the fleet can be cut. The transport manager’s instinct is no - peak days already feel tight. The FD’s instinct is yes - average utilisation is below 80%. The argument goes nowhere without operational evidence.

The transport team re-plans a set of representative weeks - a peak week, a couple of normal weeks, a quiet week - against a 28-vehicle fleet, and reads the results across them. The output:

• Cost - an annual saving of £100k-£140k depending on whether the two removed vehicles are sold or repurposed, against a mileage increase of around 3% across the rest of the fleet.
• Feasibility - three days across the sampled weeks would have needed spot-hire on 28 vehicles; modelled cost of that spot-hire is £8k.
• Vehicle utilisation - average payload utilisation rises from 72% to 81%.
• OTIF - modelled OTIF across the sampled weeks slips from 98.5% to 97.8%, the result of those three spot-hire days; if the contracts permit a 2% spot-hire fallback, OTIF holds.

The FD sees a defensible saving with the risks quantified. The transport manager sees three specific days the smaller fleet would have struggled with, and can say whether those days were genuinely peak or just anomalous. The decision gets made on the evidence, not on two conflicting instincts.

This is the pattern Filshill used to support their two-vehicle reduction (over £120,000 a year). Scenario comparison is the workflow that generalises it.

[Note: the numbers in this example are illustrative, shaped on the Filshill pattern; the bookable figures are in the published Filshill case study.]

How This Compares

Most TMS and VRP products either do not do strategic scenario modelling, or do it through a separate module that reasons over different abstractions from the operational planner. Optrak runs the scenario through the operational optimiser itself, against the same data model, the same constraints and the same heuristic catalogue as daily operations. There is no abstraction gap to defend, and no risk of the strategic model and the operational model drifting apart.

For a decision whose credibility rests on operational realism - a fleet reduction, a contract bid, a depot move - that is the difference between a number you can act on and one you have to hedge.