Fleet and Route Optimisation
A practical route plan must assign the right orders to available vehicles and drivers, sequence every stop and meet the service promise within operating and legal limits. The shortest collection of routes can fail if it ignores loading, access, time windows, driver hours or the work required at each stop.
Marketways designs routing decisions around the operation that must execute them. We combine orders, vehicles, drivers, depots, travel times, service rules and costs in one constrained model. The resulting plan states which vehicle should serve each movement, in what sequence and how dispatchers should respond when conditions change.
Define the transport decision and service rule
Routing may support parcel delivery, store replenishment, field service, employee transport, passenger movement, collection or a mixture of activities. Each operation has a different definition of a completed stop and an acceptable failure.
Marketways defines the planning horizon, dispatch frequency, service window, completion rule and escalation requirement. We also distinguish mandatory commitments from preferences. This definition determines which constraints the model must satisfy and which trade-offs management may consider.
Prepare orders as executable movements
An order record may omit the actual delivery point, access instructions, handling time, load dimensions or promised window. Duplicate or cancelled orders can consume capacity in the plan even though no movement is required.
Marketways validates addresses, coordinates, quantities, service times, dates, priorities and customer requirements. We link related pickups and deliveries and identify orders that require review. The optimiser receives movements that a dispatcher and driver can interpret.
Match each movement with a compatible vehicle
Vehicle capacity is more than weight or volume. Product temperature, hazardous classification, lift equipment, security, height restrictions, range and customer access can make an otherwise available vehicle unsuitable.
Marketways records vehicle capability, cost and availability and matches them with order requirements. Incompatible assignments remain prohibited rather than entering the objective as a cheap penalty. Management can also see which specialised requirement forces additional fleet capacity.
Represent driver availability and lawful schedules
A feasible vehicle route may still create an impossible driver schedule. Shift times, breaks, licensing, skill, overtime, starting location and hours-of-service rules determine who can execute the route.
Marketways includes these conditions in assignment and scheduling. We distinguish a vehicle that remains physically available from a driver who has exhausted the permitted duty. The model can compare relief, split shifts or another vehicle only where the organisation and applicable rules allow them.
Include depot loading and dispatch capacity
Plans often assume that every vehicle can leave the depot at its scheduled start time. In practice, docks, loaders, staging lanes, charging points and documentation can limit simultaneous departures. A route delayed before departure may miss every later window.
Marketways models loading duration, bay capacity, preparation sequence and dispatch cut-off. Routes receive start times that the depot can support. Where loading creates the constraint, the analysis shows whether changing waves, labour or staging creates more value than adding vehicles.
Use travel times that reflect departure time
Travel time between two locations can change by hour, day and direction. A route based on one average may appear feasible while placing vehicles in recurrent congestion at exactly the wrong time.
Marketways uses time-dependent travel estimates where the data and decision justify them. We preserve uncertainty rather than treating a predicted duration as certain. The route sequence and departure time can then adapt to known congestion patterns without promising precision the evidence cannot support.
Coordinate loading order with stop sequence
The route and the load plan interact. A delivery scheduled first may be physically inaccessible if later-stop products block it. Weight distribution, stacking, compartment, unloading method and pickup space can further restrict the sequence.
Marketways represents material loading rules that determine feasibility. Where detailed loading is planned separately, the routing model passes vehicle, order and sequence requirements to that process and checks the result. A route is released only when the movement and load can be executed together.
Model linked pickups, deliveries and returns
Some movements require an item to be collected before it is delivered. Others create returns, reusable packaging, cash, documents or waste that consume capacity after earlier stops. Treating routes as outbound deliveries alone can overload the vehicle or violate custody.
Marketways models precedence, pairing, evolving load and required custody. The route records what the vehicle carries after each stop. This approach keeps collection and return commitments within the same capacity and schedule decision.
Allow multiple trips and depots where operations support them
A vehicle may reload and perform another trip within the shift. It may also start or finish at a different depot when parking, inventory and driver arrangements permit. A single-trip, single-depot model can therefore overstate required fleet.
Marketways includes multi-trip or multi-depot choices only when the supporting operations exist. Reload time, inventory availability, driver handover and end-of-shift location remain explicit. The model does not create apparent savings through an operating option the organisation cannot execute.
Separate hard constraints from managed exceptions
Some conditions cannot be violated: vehicle compatibility, legal duty limits and prohibited access are examples. Other requirements may permit an authorised exception, such as a wider customer window or a subcontracted delivery.
Marketways classifies each rule and identifies who may approve an exception. Soft constraints carry a consequence that reflects the actual service or cost impact. This structure prevents the optimiser from repeatedly breaking an important promise because its numerical penalty was set too low.
Measure the complete cost of a route
Distance alone omits fixed vehicle use, driver time, overtime, tolls, energy, waiting, handling, subcontracting and failed delivery. Minimising kilometres can even increase cost if it adds a vehicle or pushes a driver beyond the shift.
Marketways builds the objective from the costs and service outcomes management needs to control. We report the components separately so that one assumed rate does not hide the reason for a recommended plan. Sensitivity analysis shows which costs materially change the fleet or route decision.
Plan for operating variation
Orders, travel times, service durations, absences and vehicle availability can change after planning. A tightly optimised schedule may have no workable response when one early stop runs late.
Marketways tests the plan under representative variation and disruption. We compare deliberate slack, standby resources, resequencing and subcontracting with the service each option protects. The result identifies where resilience is valuable instead of adding the same buffer to every route.
Replan without creating unnecessary instability
Real-time information can justify changing an active route. However, continual optimisation can confuse drivers, customers and depots, especially after loading and communication have begun. A mathematically better route may be operationally worse once the cost of change is considered.
Marketways defines the replanning interval, frozen decisions and acceptable changes. The model compares expected improvement with disruption to current execution. Dispatchers receive a stable plan unless new evidence creates enough value or risk to justify revision.
Give dispatchers approved recovery choices
A route plan cannot specify every future event. Dispatchers still need clear choices when a vehicle fails, a customer is unavailable, a stop overruns or an urgent order arrives.
Marketways converts scenario analysis into recovery rules. The rule identifies what can be reassigned, delayed, split, subcontracted or escalated and who holds the authority. This guidance allows local response without abandoning the service, safety and cost logic used to create the original plan.
Bound AI-assisted dispatch
An AI-assisted workflow can collect live vehicle, traffic, order and delivery evidence; identify an exception; run an approved optimisation; and prepare the proposed change. It can also explain which constraint caused the original plan to become infeasible.
The workflow should not ignore a safety rule, extend a driver's lawful hours, use an incompatible vehicle or alter a customer commitment without authority. Marketways tests missing locations, stale status, failed optimisation, conflicting orders, unavailable vehicles and unauthorised tool calls. Consequential exceptions remain with the responsible dispatcher or manager.
Evaluate plans against execution
A solver can reduce modelled distance while routes continue to leave late or receive frequent manual correction. The gap may arise from missing constraints, inaccurate service times, poor data or a plan that drivers cannot use.
Marketways compares planned and actual departure, sequence, travel, service, waiting, completion and cost. We also track on-time service, vehicle use, overtime, failed stops, manual changes and plan stability. These differences improve the next model and show whether optimisation changed the operating outcome.
How Marketways supports fleet and route decisions
Marketways combines Fleet and Logistics Optimisation, Operations Research and Optimisation, time-dependent travel analysis, simulation and controlled workflow design. We translate operating rules into a model, test feasibility, quantify trade-offs and design execution and recovery procedures.
The result is an explainable fleet, route and schedule plan that drivers and depots can execute, together with approved responses for the conditions most likely to disrupt it.
Related guides and evidence
References
- The vehicle-routing problem: state-of-the-art classification and review
- Thirty years of heterogeneous vehicle routing
- Vehicle-routing problems with loading constraints
- Review of dynamic vehicle routing problems
- Stochastic dynamic vehicle routing in prescriptive analytics
- Dynamic vehicle routing with random requests
- Recent advances in time-dependent vehicle routing