Learning to Run a Power Network with Trust

Winner (New Entrant Track) 4,500 USD
Winner (Anyone Track) 7,500 USD
Best Student Prize 1,000 USD

On the way towards a sustainable future and following up the success of L2RPN 2020 NeurIPS competition, this competition aims at unleashing the power of artificial intelligence even further for our real-world industrial application: controlling electricity power transmission in real-time and moving closer to truly “smart” grids using underutilized flexibilities. In 2020, participants were asked to develop an agent to be robust to unexpected events and keep delivering reliable electricity everywhere even in difficult circumstances.

In this competition, participants, while dealing with a higher penetration of renewable energy, will be asked in addition to design trustworthy agents that are able to communicate when they are in trouble, especially when they might fail. This will more concretely lead the path towards an AI assistant for human operators, who will still be responsible for managing the grid, rather than a mere blackbox agent. Join us for this Augmented Intelligence competition!

Automatic Reinforcement Learning for Dynamic Job Shop Scheduling Problem

Winners will be invited to speak at the ICAPS 2021 Competition Session.
First Prize 5,000 USD
Second Prize 2,000 USD
Third Prize 1,000 USD
Special Prizes 500 USD

In this competition, participants are invited to develop automatic reinforcement learning solutions to the dynamic job shop scheduling problem (DJSSP). The solutions are expected to automatically train promising agents on a distribution of DJSSP tasks. After a feedback phase where solutions can be developed and fine-tuned with daily feedback, those that perform best on a set of five unseen tasks win the competition.

The Flatland Challenge: Multi-Agent Reinforcement Learning on Trains

First Prize DJI Mavic Mini 2 Fly More Combo
Second Prize Oculus Quest 2
Third Prize Oculus Quest 2

This challenge tackles a key problem in the transportation world: How to efficiently manage dense traffic on complex railway networks? This is a real-world problem faced by many transportation and logistics companies around the world such as the Swiss Federal Railways and Deutsche Bahn. Your contribution may shape the way modern traffic management systems are implemented, not only in railway but also in other areas of transportation and logistics!

The Dynamic Pickup and Delivery Problem

Winners will be invited to speak at the ICAPS 2021 Competition Session.
First Prize 5,000 USD
Second Prize 3,000 USD
Third Prize 2,000 USD

The Dynamic Pickup and Delivery Problem (DPDP) is an essential problem within the logistics domain. So far, research on this problem has mainly focused on using artificial data which fails to reflect the complexity of real-world problems. With this competition, we hope to promote the research and development of techniques applicable to such problems, by building the competition around a dataset generated from real business scenarios of Huawei Technologies Ltd.

ICAPS Competitions

Read more about the rich history of competitions at ICAPS here, including more than two decades of the International Planning Competition. This year we continue this tradition and bring to you a host of challenges based on real world data with help from our partners across industry and academia. We hope this keeps you busy during the summer as we await a slightly delayed ICAPS in Fall! 🤓

Got questions? Please reach out to the ICAPS 2021 Competition Chairs.