Optimising retail logistics: the routing algorithm for Italy’s large-scale grocery retail sector

Integrating AI as a Service into Tesisquare’s TMS for optimised trip planning.

Markets

Solutions

Routing as a Service

The challenge

Large grocery retail networks face daily logistical complexity, planning intricate transport routes to hundreds of stores.

Our technology partner Tesisquare was hitting hard limits with the optimisation engine previously built into its TMS, unable to handle the strict operational constraints typical of the retail sector. This gap forced planners to break daily planning into numerous separate runs – up to 14 executions per instance – and to make extensive manual adjustments to the trips the system generated. The process could take up to half a working day, resulting in significant inefficiencies, difficulty meeting time windows, and a final solution that fell short on fleet utilisation.

The challenge was to fully automate the process, while keeping it completely aligned with real operational needs on the ground.

How we create value

To overcome these operational inefficiencies, Optit provided a standard routing service for retail, natively integrated as a “black-box” solution (AI as a Service) within Tesisquare’s TMS via a REST API.

Without adding any new digital tools for end users, the engine runs on an advanced algorithm that can handle both “green-field” planning of entire order sets and the synchronous recalculation of individual trips.

The model natively accounts for complex constraints: vehicle reuse across multiple trips and depots (multi-trip), hard and soft time windows, order-vehicle compatibility (such as temperature control), multi-dimensional capacity (pallets, kilograms and cubic metres), and statutory rules on drivers’ hours.

Beyond calculating optimal routes, the system returns a detailed set of operational KPIs and automatic warnings on constraint violations, giving planners immediate decision support directly within the TMS.

Impact

Adopting Optit’s solutions has radically transformed supply chain performance:

A shift from fragmented workflows to a single automatic run per planning area.
A sharp cut in manual adjustments to trips, consistently below the 15% threshold.
A major reduction in planning time, freeing up hours for higher-value work.
Improved logistics efficiency, driven by an increase in average fleet utilisation and a reduction in total kilometres