3rd International Reinforcement Learning Bootcamp 2026

16 Sept 2026, 11:00 → 18 Sept 2026, 17:00 Europe/Vienna

Unipark Nonntal (HS Agnes Muthspiel) (PLUS University Salzburg)

Simon Hirlaender, Olga Mironova

This bootcamp begins with fundamental principles and systematically progresses to advanced topics in Reinforcement Learning, providing a structured and in-depth educational experience.

Participants will attend lectures, interactive workshops, and practical coding sessions delivered by leading experts. The bootcamp is designed for students, industry professionals, and researchers who have acquired a foundational understanding of machine learning and seek to deepen their expertise in Reinforcement Learning.

The 3rd International Reinforcement Learning Bootcamp takes place September 16–18, 2026, at PLUS University Salzburg.

Full program, speakers, and details: RL Bootcamp - Reinforcement Learning

Wednesday 16 September

11:00 → 12:00 Arrival, Registration & Welcome Coffee

12:00 → 13:00 Opening Lunch (ARGE Beisl) (self-paid)

13:00 → 13:15 Opening & Orientation

Scope of Bootcamp 2026, schedule, logistics, and onboarding.

Speaker: Simon Hirlaender (Paris Lodron Universität Salzburg (PLUS) / IDA Lab)

RL_Bootcamp_2026_Opening_and_Orientation.pdf

13:15 → 14:00 Lecture 1: Introduction to Reinforcement Learning: "No Math, No Panic"

RL mindset, agent-environment loop, return G_t, discount factor gamma, and exploration vs exploitation.

Speakers: Olga Mironova (Paris Lodron Universität Salzburg (PLUS) / IDA Lab), Simon Hirlaender (Paris Lodron Universität Salzburg (PLUS) / IDA Lab)

RL_Bootcamp_2026_Lecture_1_No_Math_No_Panic.pdf

14:00 → 15:00 Lecture 2: Reinforcement Learning Fundamentals: "From Words to Equations"

Markov Decision Processes, Bellman expectation and optimality equations, value decomposition (V, Q, A), and tabular updates.

Speaker: Simon Hirlaender (Paris Lodron Universität Salzburg (PLUS) / IDA Lab)

RL_Bootcamp_2026_Lecture_2_From_Words_to_Equations.pdf

15:00 → 15:30 Coffee Break 1

15:30 → 17:30 Tutorial 1: Tabular RL & Discrete MDPs

Interactive Google Colab Maze environment; implementing Bellman updates, comparing SARSA vs Q-Learning.

Speakers: Leander Grech (Paris Lodron Universität Salzburg (PLUS)), Léa Keller (Paris Lodron Universität Salzburg (PLUS))

RL_Bootcamp_2026_Tutorial_1_Tabular_RL.pdf

17:30 → 18:15 Keynote 1: Designing for Intelligence: Particle Accelerators in the AI Era

Real-time beam steering, AWAKE electron beam stabilization, and safety constraints at CERN.

Speaker: Verena Kain (CERN, Switzerland)

19:00 → 21:00 Welcome Dinner (ARGE Beisl) (self-paid)

Thursday 17 September

08:30 → 09:00 Morning Desk & Speaker Reception

09:00 → 09:45 Lecture 3: Policy Gradients and Actor Critics

Continuous state spaces, policy parameterization pi_theta(a|s), REINFORCE, Advantage Actor-Critic, and PPO.

Speaker: Thomas Gallien (JOANNEUM RESEARCH / AI Austria)

RL_Bootcamp_2026_Lecture_3_Policy_Gradients.pdf

09:45 → 10:30 Keynote 2: What Should Models Model in Model-Based RL?

Continual learning, game-theoretic foundations, world models, predictive representations.

Speaker: Michael Bowling (University of Alberta / Amii, Canada)

Keynote_2_Michael_Bowling_Model_Based_RL.pdf

10:30 → 11:00 Coffee Break 2

11:00 → 11:45 Keynote 3: The Benefits of Satisficing in Reinforcement Learning

MDP regret bounds, PAC learning, satisficing sequential decision making.

Speaker: Ronald Ortner (Montanuniversität Leoben, Austria)

RL_Bootcamp_2026_Keynote_3_Ronald_Ortner.pdf

11:45 → 12:30 Keynote 4: Learning Humanoid Robot Policies for Industrial Tasks

Stochastic optimal control, dynamic movement primitives, sim-to-real transfer, humanoid robotics.

Speaker: Elmar Rückert (Montanuniversität Leoben - CPS, Austria)

RL_Bootcamp_2026_Keynote_4_Elmar_Rueckert.pdf

12:30 → 12:45 Official Group Photograph (Pia Fronia)

12:45 → 14:00 Lunch Break (ARGE Beisl) (self-paid)

14:00 → 15:30 Tutorial 2: Air Traffic Control as an MDP (Intermediate)

MDP formulation for air traffic management, multi-aircraft continuous state spaces, first trained agent vs baseline in Flight4DEnv.

Speaker: Leander Grech (Paris Lodron Universität Salzburg (PLUS))

RL_Bootcamp_2026_Tutorial_2_Continuous_Control_Reality_Gap.pdf

15:30 → 16:00 Coffee Break 3

16:00 → 18:00 Tutorial 3: The Flight Challenge: How to Compete (Advanced Tournament)

Air Traffic Control Tournament in Flight4DEnv; custom rewards, safe separation minimums, live leaderboard.

Speaker: Leander Grech (Paris Lodron Universität Salzburg (PLUS))

RL_Bootcamp_2026_Tutorial_3_Flight_Challenge.pdf

19:00 → 21:00 Official Conference Social Dinner (Dai Pai) (self-paid)

Friday 18 September

09:00 → 10:30 Interactive Wrap-Up Tutorial: Findings & Expert Panel

Flight Challenge championship debrief, strategy evaluation, open audience Q&A with keynote speakers.

Speakers: Leander Grech (Paris Lodron Universität Salzburg (PLUS)), Olga Mironova (Paris Lodron Universität Salzburg (PLUS) / IDA Lab), Simon Hirlaender (Paris Lodron Universität Salzburg (PLUS) / IDA Lab), Thomas Gallien (JOANNEUM RESEARCH / AI Austria)

10:30 → 11:00 Coffee Break 4

11:00 → 12:00 Lecture 4: Agentic AI in the Context of Reinforcement Learning

Token-level MDPs, verifiable execution loops, LLM agents with RL environments, tool use and reflection. Includes 5-min Lightning Pitch: DLR LUMEN Rocket Propulsion RL Benchmark (Jonas Dauer, DLR Lampoldshausen).

Speakers: Jonas Dauer (German Aerospace Center (DLR)), Marius-Constantin Dinu (ExtensityAI / AI Austria)

RL_Bootcamp_2026_Lecture_4_Agentic_AI.pdf

12:00 → 13:00 Tournament Final, Awards Ceremony & Closing Remarks

Announcement of winners (3x Sutton & Barto textbook prizes), certificate distribution, and Bootcamp 2027 outlook.

Speakers: Leander Grech (Paris Lodron Universität Salzburg (PLUS)), Olga Mironova (Paris Lodron Universität Salzburg (PLUS) / IDA Lab), Simon Hirlaender (Paris Lodron Universität Salzburg (PLUS) / IDA Lab)

RL_Bootcamp_2026_Closing.pdf

13:00 → 14:00 Farewell Lunch (ARGE Beisl) (self-paid)

15:00 → 16:00 Salzburg Cultural Walking Tour (Kang Yuan Hau)