Workshops

As part of the Summer School 2026, we will offer several hands-on workshops in which you will be able to dive deeper into different neuroengineering topics and directly apply that knowledge by working on the proposed challenges in small teams!   

Each workshop will begin with an introductory session. On Wednesday, you will have all day to work in your assigned challenge and lastly, on Thursday, your team will present their results to the rest of the Summer School audience. 

Please, read each workshop description carefully and then fill in this formular to express your interest in each of the workshops. We will use this information to do the workshop assignment and you will be contacted soon with your assigned workshop.

Find the full workshop descriptions below.

Simon Jacob - Translational NeuroTechnology Laboratory, TUM

From Neural Activity to Action Choice: Decoding Behavior in the Rodent Brain

Philipp Beckerle - Chair for Autonomous Systems and Mechatronics, FAU

Muscle Synergy Extraction, Analysis, and Translation for Neuroengineering Applications

Simon Baumann - Laboratory of Integrative Neuroscience and Cognition, Georgetown University

Analysis challenge on multielectrode Neuropixels data from the auditory-motor network of nonhuman primates.

Dario Martelli -  Gait and Motion Analysis Lab, MedStar Health

Interactive Neuroengineering: Decoding EMG Signals for Real-Time Human Control 

Joachim Hermsdörfer - Chair of Human Movement Science, TUM

ADL performance: From degrees of freedom to robust quantification 

Workshop Descriptions

Simon Jacob - Translational NeuroTechnology Laboratory, TUM

Title: 

From Neural Activity to Action Choice: Decoding Behavior in the Rodent Brain

Description:

How does the brain convert sensory information into action? In this workshop, participants will analyze large-scale neural recordings in the prefrontal cortex from rodents performing a decision task to reconstruct the steps linking external cues to behavioral choice. Using a guided, hands-on pipeline, students will learn how raw population activity is processed, aligned to task events, and transformed into interpretable measures of neural coding. They will then examine how individual neurons and neural populations represent task variables, and test whether ongoing activity can be used to decode the animal’s upcoming action. The project is designed as a collaborative analysis challenge rather than a programming competition, so students with different levels of experience can contribute to the same scientific question while building intuition for modern neural data analysis. By integrating preprocessing, single-neuron selectivity, and population decoding into a coherent workflow, the workshop introduces core concepts in neuroengineering through a concrete problem in systems neuroscience: reading out action-related information from large-scale brain recordings.

This Workshop will take place at the Klinikum Rechts der Isar

Philipp Beckerle - Chair for Autonomous Systems and Mechatronics, FAU

Title:

Muscle Synergy Extraction, Analysis, and Translation for Neuroengineering Applications

Description:

This workshop introduces participants to the concept of muscle synergies. The workshop explores the hypothesis that the Central Nervous System achieves efficient and adaptable movement generation through low-dimensional modular control strategies, referred to as muscle synergies, reducing the complexity and energetic cost of coordinating the musculoskeletal system. Understanding these mechanisms is valuable not only for studying human motor control, but also for assessing neurological impairments, such as those caused by stroke, and for developing more adaptive and human-centered assistive technologies.

Participants will be introduced to the Delsys Trigno Avanti system, including both surface electromyography (sEMG) and inertial measurement unit (IMU) recordings, as well as basic data acquisition and preprocessing techniques for capturing and analyzing human movement and muscle activity. Using these signals, students will then extract and analyze muscle synergies through dimensionality reduction techniques such as Non-negative Matrix Factorization (NMF).

Students will be able to explore real-time visualization of movement and synergy activations, as well as basic applications of synergy-based control in virtual models, myoelectric exoskeletons, and neurorehabilitation. 

Throughout the session, students will combine concepts from biomechanics, signal processing, and computational motor control in a practical guided workflow including:

• EMG acquisition and preprocessing

• Muscle synergy extraction

• Synergy analysis and interpretation

• Real-time visualization and interaction using computational models and interactive platforms

This workshop will take place at the Institute for Cognitive Systems

Simon Baumann - Laboratory of Integrative Neuroscience and Cognition, Georgetown University

Title:

Analysis challenge on multielectrode Neuropixels data from the auditory-motor network of nonhuman primates.

Description:

Next-generation silicon-based probes such as Neuropixels provide a game-changing opportunity to simultaneously obtain neuronal data from hundreds to thousands of neurons, potentially from several brain areas of a neural network. However, these opportunities also come with challenges: how to deal efficiently with massive amounts of data (computation-wise, storage etc.) and how to separate the wheat from the chaff, the neurons and time periods that provide the most information for the research questions. In this analysis challenge, we will provide original Neuropixels data from several brain areas of the auditory-motor network of nonhuman primates (nhps) while they performed a task on a 'monkey piano'.

In an initial session, we will discuss the general properties of the data, specific challenges that come with the nature of the data and possible directions for solutions. Then we will hand out examples of data from different brain areas to each group (of ideally 2-3 but individual work possible). The provided data will contain raw data, task timing information and, for efficiency reasons, we will provide sorted spike times from Kilosort 3. Your task will be to find solutions to the analysis challenges, find neurons that respond to the task, and if time is available you are free to set your own challenges e.g. create recording depths maps, neuronal interactions etc. You will get bonus points if you figure out from which of a limited number of brain areas your data has originally been sampled. On the last day, you will present your solutions, discuss the challenges and directions for future improvements.

Experience with the analysis of neurophysiological data is useful but not essential. Each group will need access to a computer with ideally +32 GB working memory, ~200 GB hard disk and ideally a decent graphics card. In case of specific performance bottlenecks, pre-processed data can be provided. Additional computers with lower specifications for each group member for less computation intensive tasks will be useful. Each computer should have either Matlab or an analysis framework of your choice (Python etc.) installed. Basic programming skills (in Matlab or your software of choice) are expected. 

This workshop will take place at the Institute for Cognitive Systems.

Dario Martelli -  Gait and Motion Analysis Lab, MedStar Health

Title:

Interactive Neuroengineering: Decoding EMG Signals for Real-Time Human Control 

Description:

This hands-on workshop focuses on building systems that translate muscle activity into real-time control signals for interactive applications. Students will use electromyography (EMG) data to drive interactive applications or simple decision-based interfaces. By the end of the workshop, participants will have built a working system that converts human muscle activity into real-time interaction, illustrating core concepts used in neuroprosthetics, rehabilitation engineering, and human–machine interfaces. 

This workshop will take place at the Institute for Cognitive Systems

Joachim Hermsdörfer - Chair of Human Movement Science, TUM

Title:

ADL performance: From degrees of freedom to robust quantification

Description:

Kinematic analysis of ADL performance:

• Choice of task and team assignment

• Pipeline from recording to raw data

• Dealing with noise and gaps

• Feature extraction

The association with capacity:

• Clinical assessments, i.a. BB, Grip, Tap

• Sequence effects and targets

• Association with submaximal performance (ADL)

Sensitivity of ADL performance:

• Impede cognitive resources (e.g., dual tasking)

• Impede sensation (e.g., rubber gloves, wearing strong glasses)

• Impede strength (e.g., added weights, prior exertion)

A.I. support (if time allows):

• Can A.I. detect content errors?

• Is markerless tracking reliable?

• How well does A.I. analyze complex time-series?

This workshop will take place at the Chair of Human Movement Science's Living Lab, at the TUM Campus in the Olympic Park.