MaGIC 2026

MaGIC 2026 is a three-day event bringing together researchers and practitioners from industry, academia, and national labs to share advances in modeling, simulation, and machine-ground interaction.

Date: September 22–24, 2026.

Location: DeLuca Forum, Discovery Building, University of Wisconsin-Madison — 330 N Orchard St, Madison, WI 53715.

Organizers: Dan Negrut and Radu Serban, University of Wisconsin-Madison.

Funding Support: National Science Foundation.

Registration Link: https://charge.wisc.edu/GraingerInstitute/magic.

Registration Fee: $150 for industry participants; free for academia, state or federal employees, and consortium members.

At its 14th edition, the 2026 MaGIC meeting emphasizes modeling and simulation in fields such as terramechanics, mechatronics, bio-robotics, geomechanics, and embodied AI. The event is focused equally on terrestrial and extraterrestrial applications. This is an informal gathering that seeks to bring together individuals from industry, academia, and research labs in a collaborative “let’s learn from each other” environment. The goal is to facilitate technology transfer and promote cross-pollination between industry and academia within a pre-competitive setting.

Overview of the program:
Training Day: Tuesday, September 22, 2026 — Chrono tutorials, for those looking to use the Chrono simulator in their work.
Invited Talks Day 1: Wednesday, September 23, 2026 — invited talks in the morning and afternoon. Student poster session in the early evening.
Invited Talks Day 2: Thursday, September 24, 2026 — invited talks in the morning and afternoon.

NOTE: This event is free for participants from (1) state or federal institutions, (2) companies that are members of the Machine-Ground Interaction Consortium, and (3) academia. Otherwise, the registration is $150/person. Contact Dan Negrut at negrut@wisc.edu if you are a participant from a small company who can’t afford the registration fee.

List of speakers (still in flux, listed in alphabetical order):

September 22, 2026: Tutorials highlighting the use of Project Chrono

As part of MaGIC 2026, this tutorial introduces Project Chrono, an open-source multi-physics simulation framework. Through a series of hands-on sessions, participants will explore fluid-solid interaction, terramechanics, Discrete Element Method (DEM) simulation, sensor simulation, reinforcement learning via the Gymnasium environment, and core Chrono modeling and simulation workflows. There will be two parallel tracks of the tutorial sessions; all sessions will be recorded and made available to the participants after the conclusion of the event.

Overview of the tutorial sessions (topics and duration still in flux)

Chrono DEM Solver

Introduction to Chrono’s GPU-based Discrete Element Method (DEM) solver for simulating granular materials.

Duration: 3 hours.

Chrono FSI framework for fluid-solid interaction

Introduction to the Fluid Solid Interaction framework in Chrono that allows simulating interaction between mechanisms containing rigid and flexible bodies with water and granular material. We will go through the basics of how the framework is set up and how information is exchanged between different parts of Chrono. We will then walk through a hands-on example leveraging the framework (see picture for simulation types supported by this infrastructure).

Duration: 2 hours.

Chrono CRM, high-fidelity terramechanics model

Introduction to a Chrono terramechanics model that regards the terrain as a continuum, instead of treating it as a collection of fine particles/elements. Chrono’s Continuum Representation Model (CRM) is physics-based, runs fast, and produces accurate results. We will cover the theory and the numerical method behind Chrono CRM and then walk through a demonstration of setting up a drawbar pull test with a full scale rover. The walk-through will be hands-on where we will live code the simulation setup, build it, run it, and visualize the results.

Duration: 2.5 hours.

Synchrono, distributed co-simulation capabilities

Synchrono is a co-simulation framework that enables scalable, distributed simulation of robotic, vehicular, and infrastructure systems. It allows multiple Chrono-based simulation nodes—each representing a subsystem like a vehicle, sensor, or infrastructure component—to run concurrently on different processes or machines while maintaining synchronized state. In this session, we will cover the architecture of Chrono::Synchrono, the message-passing system, the data-distribution system used for inter-node communication, and how the framework supports modular simulation design across multiple computing nodes. We will then explore two example scenarios:

• Multiple autonomous vehicles driving as a convoy, with each vehicle simulated as an independent agent using its sensors and control logic.
• Multiple rovers traversing on SCM deformable terrain.

Duration: 1 hour.

Off-Road Autonomy Design on Chrono CRM Terrain

Application 1: RL Training with Terrain-Aware Domain Randomization

• Train a quadruped robot using reinforcement learning across various terrain types supported by Chrono.
• Conduct 95% of training on rigid terrain; apply the remaining 5% on CRM granular terrain for fine-tuning.
• Highlight performance gains from terrain-wise domain randomization and show generalization benefits on challenging off-road environments.

Application 2: Learning-Based Reduced Order Modeling (ROM) for Soil-Tool Interaction

• Use data-driven world model learning to capture the dynamics of a blade interacting with granular soil.
• Develop a simplified model (ROM) that approximates the soil-blade interaction with reduced computational cost.
• Discuss how the learned model can support downstream tasks such as motion planning and closed-loop control.

Duration: 2 hours.

Chrono::Sensor for software-in-the-loop (SITL) development

Chrono::Sensor is a framework that simulates common sensors (i.e., camera, IMU, GPS, LiDAR) operating in Chrono virtual worlds. This allows for fast software-in-the-loop development of robotic and ground vehicle systems. We will begin with an introduction to the Chrono::Sensor API and implementation (manager, sensor filters). We will then learn to integrate the built in GPS, IMU, Camera and LiDAR models into a Chrono::Vehicle simulation. We will then briefly cover how to extend this system for SITL using Chrono::ROS integration and simulating a robot using an Extended Kalman Filter for perception. Finally, we cover camera models available and their use cases.

Chrono Gymnasium Environment

An introduction to the Chrono Gymnasium environment, a framework that wraps Chrono simulations as standard OpenAI Gymnasium-compatible environments for use in reinforcement learning, optimization, and calibration workflows. This session covers the structure of the environment, including how observations, actions, and reward signals are defined, and walks through how to configure and use it for custom simulation scenarios. We will then demonstrate two practical applications:

• Bayesian Optimization: Using the Gymnasium environment to efficiently optimize simulation parameters or control policies through surrogate-model-based search.
• Bayesian Calibration: Leveraging the environment to calibrate simulation parameters against real-world data using probabilistic inference techniques.

Duration: 2 hours.