Abstract: 

The current era is driven by eCommerce platforms and the COVID pandemic gave rise to the need for home delivery. The end consumers have multiple options to cater for their needs and in that case, the eCommerce platforms have to provide on-time and quality delivery to stay ahead in the market and at the same time boost their profit margins.

Route Optimization is one of the most critical aspects of planning and transportation as it ensures that deliveries always arrive on time and carry out the same with the lowest possible cost and energy consumption. However, there are a lot of variables that eCommerce platforms need to consider in a real-time scenario.

During this unfortunate COVID pandemic, eCommerce platforms are dealing with a huge inflow of e-commerce orders from a variety of customers scattered throughout a city, country or even across the globe. This gives rise to a humongous number of variables come into play that simply cannot be solved using conventional methods in a practical amount of time. With the recent developments in AI, machine learning and cloud data the entire game of route optimization has begun to change. AI is continuously retrieving data, learning from it, and searching for improved methods to ensure the most optimal routes for the drivers.

In the novel solution, we are trying to solve the multi-objective vehicle routing problem with optimization variables like minimizing the cost to deliver, the number of vehicles and delivery time.

Outline/Structure of the Talk

1. Introduction of Route Optimisation.
2. Classical approaches to Multi-Vehicle route optimization and its limitations.
3. Introduction to key concepts of
Multi-objective Optimization
Deep Reinforcement Learning
Genetic Algorithms and other Metaheuristics
4. Combination of Deep RL-based algorithm with a metaheuristic approach.
5. Training methodology, use-case and result discussion.
6. Other Business Aspects

Learning Outcome
1.The NP-Hard problem of Route Optimisation.
Gain an understanding of deep reinforcement learning and metaheuristics based route
optimization techniques.
2.How the combination of both Reinforcement Learning and Metaheuristics together can be
used to produce a solution for different cases of route optimisation.
3. How to build a production grade ML-pipeline for route optimisation

Target Audience
Data Scientist and Machine Learning Engineer, Data Engineers, Data Architects

Bio: 

Sushant has a bachelor’s in Materials Engineering from the Indian Institute of Technology, Kharagpur. He has extensive work experience in Reinforcement Learning, Computer Vision with AR technologies and created end-to-end pipelines and data products from conceptualisation to deployment phase for various engagements. He is currently working as a Senior Data Scientist at Publicis Sapient.