Route optimisation for Tesco
We co-created a last mile delivery system with Tesco, optimising vehicle schedules throughout the UK - saving them millions of miles, reducing costs and their carbon footprint.
Challenge
As online deliveries became increasingly popular, Tesco recognised the need to develop technology that improved their online proposition, and cemented their position as market leaders. Their existing vehicle scheduling system couldn’t be adapted to fulfil their needs and so, having gone to market, they selected Satalia as partners to build a bespoke vehicle scheduling system.
Outcomes
8%
decrease in fuel consumption per order
11.2
million miles saved
5%
increase in the number of drops per hour
Same Day
enabled same-day delivery options
Outcomes
8%
decrease in fuel consumption per order
11.2
million miles saved
5%
increase in the number of drops per hour
Same Day
enabled same-day delivery options
Solution
Working with Tesco, we combined machine learning, optimisation and software engineering to build a market leading vehicle routing and scheduling system. Its objective was to improve both operational efficiency and the customer experience, and it did so whilst accounting for a wide range of unique business rules, including vehicle types and capacity, shift patterns, order weight, lunch breaks and road speeds.
As a customer requests delivery slots, it analyses their basket size, driver availability and vehicle capacity. Then, having calculated travel time from their location to every other location in the schedule, the algorithm shows only the feasible slots back to the customer. This happens in less than a second. These dynamic, constantly updated schedules mean slots are offered in accordance to the actual, accurate and real-time capacity of the fleet. No over promising slots that can’t be made. No limiting orders due to lack of operational visibility.
The process
The system was developed jointly by a team with experts from both Tesco and Satalia. Using agile methodologies, and through continuous collaboration, we were able to fulfil 100% of their requirements.
The key stages of the project were as follows:
Discovery: The project team worked across operational units to gather their requirements and objectives, as well as understand Tesco existing data and infrastructure.
Proof of concept: Satalia built a thinned down version of the solution, tested on a subset of Tesco’s data, and iterated to ensure peak performance.
Pilot: The project team ran a pilot across a single catchment, measured and improved performance, and adapted to ensure the system was production-ready.
Rollout: The system was rolled out across Tesco’s entire estate. Satalia then focused on knowledge share and upskilling to ensure the system could be effectively handed over to the Tesco team.
What happened next?
Tesco continue to build upon the system they worked with Satalia to create, and amongst other initiatives, have since implemented dynamic feedback loops to ensure continuous improvement to their delivery offering. Both the system and their relationship with Satalia continues to go from strength to strength.
Having developed a bespoke system with Tesco, we re-built a highly customised vehicle routing system to serve the rest of the market. Our system has an updated set of algorithms, includes a set of advanced machine learning features,and works across the last-mile, middle-mile and service management sectors.