Reinforcement Learning: Implementation – Designing and Building RL Environments
Hands-on workshop on translating real-world problems into reinforcement learning tasks and designing, implementing, and evaluating custom RL environments using Python and Gym.
How can a real-world problem be translated into a task that an AI agent can learn to solve? The workshop “Reinforcement Learning: Implementation – Designing and Building RL Environments,” organized by the AI Service Center Berlin-Brandenburg at the Hasso Plattner Institute, provides a practical introduction to designing and implementing reinforcement learning environments.
Using a simple application such as controlling an energy storage system, participants work through the process of formulating a reinforcement learning task. They learn how to determine whether RL is an appropriate approach, define states, possible actions, and reward signals, and implement a corresponding environment using the Python library Gym.
The workshop also covers how to evaluate and refine an RL setup, helping participants understand how to apply the development process to their own challenges. It is designed for participants who already have basic knowledge of Python and reinforcement learning and want to move from theoretical concepts to practical implementation.
- Date: December 16, 2026
- Time: 10:00 AM–1:00 PM
- Location: Hasso Plattner Institute, Prof.-Dr.-Helmert-Straße 2–3, 14482 Potsdam
- Organizer: AI Service Center Berlin-Brandenburg at HPI
- Admission: Free, binding registration via Eventbrite
- Language: German
- Audience: Developers, students, researchers, and AI practitioners with basic knowledge of Python and reinforcement learning
- Focus: Reinforcement learning, RL environment design, agent-environment interaction, states, actions, reward signals, and Python
- Practical component: Designing and implementing an RL environment with the Python library Gym, using an energy storage control task as a possible example
- Learning objectives: Assess the suitability of RL, formulate tasks, implement environments, and evaluate and adapt approaches
- Prerequisites: Basic Python and reinforcement learning knowledge; familiarity with agents, states, actions, rewards, and policies
- What to bring: A laptop
- Note: Photo, audio, and video recordings will be made during the event