Route Planning & Optimization
Route planning and optimization is the engine that decides, for a given set of orders, which stops go on which vehicle, in what sequence, and by what path. Done well, it turns a pile of loose delivery addresses into a tight schedule that fits the fleet's capacity and the day's time windows; done poorly, it means half-empty trucks, drivers backtracking across town, and missed delivery appointments.
At its mathematical core, route optimization is a variant of the vehicle routing problem (VRP) — a well-studied but computationally hard class of problem where the "best" answer for more than a handful of stops can't be found by brute force in any reasonable time. Real-world TMS route engines use heuristics and metaheuristics (algorithms that find a very good, though not always mathematically perfect, solution quickly) layered with hard business constraints: vehicle capacity by weight and volume, driver shift limits, customer-specific delivery windows, and mandatory stop sequences (for example, cash-on-delivery stops that must be visited before certain other stops).
The output isn't just a sequence of addresses — a usable route plan also accounts for realistic drive times between stops (not straight-line distance), service time at each stop, and traffic patterns by time of day.
Most B2B and appointment-based deliveries come with a window — a store might only accept receiving between 7 and 9 a.m., a construction site might close at 3 p.m. A route optimizer treats these windows as hard constraints and will restructure the entire route rather than violate one, sometimes at the cost of extra mileage elsewhere. Soft constraints — customer preference for morning delivery, for instance — are typically weighted rather than enforced absolutely, letting the planner trade off preference against efficiency.
Static route planning builds the day's routes the night before or early morning, based on known orders. Dynamic (or real-time) route planning continuously re-optimizes as new orders come in, a driver runs late, or a vehicle breaks down — inserting new stops into existing routes or reshuffling the whole plan. Dynamic planning matters most in same-day delivery and field service operations where the order book keeps changing after the day has already started.
The mathematically optimal route on paper sometimes conflicts with how drivers actually work — a route that's 5% shorter in distance but unfamiliar to the driver may take longer in practice, and drivers who repeatedly get reassigned to different routes lose the local knowledge that makes them fast. Many operations therefore favor "dedicated" or "master" routes — a driver serving broadly the same territory day after day — with the optimizer making incremental adjustments rather than rebuilding the plan from scratch every morning.
Common KPIs used to judge a route plan include stops per hour, miles (or kilometers) per stop, on-time delivery rate, and vehicle fill rate (how much of available capacity is actually used). Tracking these over time — not just for a single day's plan — reveals whether the optimizer's assumptions (drive time estimates, service time per stop) match reality, and whether they need retuning.