Route Optimisation Model

The story

The client, a unicorn software company in Sweden, operates in 34 countries, connecting with 100+ print partners and serving tens of thousands of API customers, including Canva.com, OptimalPrint, Mapiful, Shopify, and Etsy stores. Their focus on local production aims to make it faster, smarter, and greener. Their products, On Demand Print and Optimal Print, aim to ensure that global customers receive orders within 72 hours. They sought to optimize delivery routes for vehicles with limited capacity to ensure timely and efficient delivery of customized print products. To achieve operational excellence, they needed highly skilled data scientists and ML engineers to build scalable ML models for optimal route selection, maximizing delivery efficiency.

The challenge

The client faced the challenge of finding the best delivery routes for vehicles with limited carrying capacity. To meet this challenge, they needed a solution that could estimate package weights, demand, vehicle requirements, moving distances, and customer locations while ensuring that delivery costs were minimized and vehicle capacity was not exceeded.

The solution

Our team provided a comprehensive solution to the client’s challenge, which involved addressing the Capacitated Vehicle Routing Problem. Using Python, Solver Pulp, and Graph theory, we conducted data analysis to estimate package weights, demand, vehicle requirements, moving distances, and customer locations. To enhance visualization, we harnessed the Google Maps API and Matplotlib libraries in Python to map customer locations on Google Maps. Ensuring data security and scalability, our experts employed the S3 bucket framework. Our team, consisting of over five members, including data scientists, utilized a tech stack that included Python, Solver Pulp, Graph theory, Matplotlib, and the Google Maps API. The deliverables included an ML model for optimal route selection, a solution for the routing problem, and the implementation of robust data security and scalability through the S3 bucket framework.

The outcome

We successfully delivered a highly accurate ML model that optimized vehicle requirements and moving distances for the fleet, resulting in a remarkable cost reduction of over 60%, thus enhancing delivery efficiency and cost-effectiveness. Furthermore, our team’s expertise in data science contributed to the client achieving operational excellence, ensuring timely deliveries and product quality. Through the application of Python, Solver Pulp, Graph theory, Matplotlib, and the Google Maps API, we crafted a solution that not only met but exceeded the client’s objectives, facilitating efficient product delivery to customers across countries.

Planning to outsource software development services?
Contact sales, to start a project, now.