Max Planck Institute for Intelligent Systems

• 1. Introduction

  ›

  On the 18th of March 2011, a research institute opened its doors across two cities in Germany, and its founding question was deceptively simple: what, exactly, is intelligence? The Max Planck Institute for Intelligent Systems, known as MPI-IS, planted itself in both Stuttgart and Tübingen, splitting its energies between two campuses with a single shared ambition. The scientists there wanted to understand how intelligent systems perceive, learn, and interact with the world. Then they wanted to use that understanding to build something that actually benefits people.

  MPI-IS is one of 84 institutes within the Max Planck Society, which puts it inside one of the most storied research networks in Europe. But what sets this particular institute apart is its refusal to stay in a single lane. Its researchers work across hardware, software, and theory at once, drawing on computational, physical, and social perspectives together. The questions ahead are hard ones: how do machines learn robust behaviors, how does robotic touch work, and how should artificial intelligence serve society rather than undermine it?

• 2. Six Departments, One Vision

  ›

  Bernhard Schölkopf leads the Empirical Inference department in Tübingen, a group focused on the statistical and mathematical underpinnings of machine learning. Across the two campuses, five other departments anchor the institute's core scientific mission. Katherine Kuchenbecker directs Haptic Intelligence in Stuttgart, studying how machines sense and respond to touch. Michael J. Black heads Perceiving Systems in Tübingen, where the focus is on how machines interpret the physical world visually.

  In Stuttgart, Christoph Keplinger leads Robotic Materials and Physical Intelligence, a department that sits at the intersection of materials science and robotics. Moritz Hardt directs the Social Foundations of Computation in Tübingen, which may be the most unusual department of all: it treats the social consequences of computation as a primary research object, not an afterthought. These six departments together define the institute's range, from the tactile to the statistical, from the robotic to the societal.

  Organization at MPI-IS runs through a faculty of Directors, Group Leaders, and Max Planck Fellows, each guiding a different slice of research. That structure allows the institute to hold many lines of inquiry simultaneously without any single department dominating the others.

• 3. Research Groups at the Frontier

  ›

  Beyond the six departments, MPI-IS supports a dense network of independent research groups, each led by a specialist pursuing a narrowly defined problem. Maksym Andriushchenko runs the AI Safety and Alignment group in Tübingen, and Sahar Abdelnabi leads Cooperative Machine Intelligence for People-Aligned Safe Systems, also in Tübingen. The names alone signal the institute's preoccupation: alignment, safety, and social consequence are recurring themes across multiple independent teams.

  In Stuttgart, Antonia Georgopoulou directs Cyborg Robotics and Intelligent Sensing, while Florian Hartmann leads Biomimetic Materials and Machines, designing mechanical systems that take cues from biological ones. Philipp Müller's Embodied Social Interaction group also sits in Stuttgart, probing how machines can participate in human social contexts. Janneke Schwane studies the Neuromechanics of Movement, and Buse Aktaş leads Robotic Composites and Compositions, both based in Stuttgart as well.

  On the Tübingen side, Jonas Geiping runs Safety- and Efficiency-Aligned Learning, and Wieland Brendel leads Robust Machine Learning. Shiwei Liu heads the Wild, Efficient, and Innovative AI Lab, while Maximilian Dax directs Science and Probabilistic Intelligence. Ksenia Keplinger in Stuttgart leads Organizational Leadership and Diversity, a group whose focus on the human side of research institutions has few direct parallels elsewhere in machine learning.

  Two Max Planck Fellow Groups extend the institute's reach further still: Achim Menges leads Intelligent Construction and Building Systems, and Andreas Krause runs Interactive Learning in collaboration with ETH Zurich.

• 4. The Learning Systems Partnership

  ›

  MPI-IS and ETH Zurich built something rare in European academia when they launched the Max Planck ETH Center for Learning Systems, known as CLS. It stands as the first joint doctoral program ever created between ETH Zurich and the Max Planck Society, a distinction that marks it as a structural experiment as much as a scientific one. The program admitted its first fellows in 2015 and has since grown to include 112 doctoral students and post-docs as Fellows or Associated Fellows.

  Fifty directors, professors, and research group leaders count themselves as members or associated members of CLS today. The program's governance runs jointly, pulling together supervisors from both institutions around doctoral candidates who benefit from both networks at once. In July 2019, the Max Planck Society and ETH Zurich agreed to extend the program's funding through 2025, a signal that both sides considered the experiment worth sustaining. The CLS model suggests a possible template for cross-institutional doctoral training in fields that move faster than any single university can track.

• 5. Cyber Valley and Industry

  ›

  Since December 2016, MPI-IS has been part of Cyber Valley, a research network built around the region of Baden-Württemberg. The partners span public institutions and private industry: the University of Stuttgart, the University of Tübingen, and the Fraunhofer Gesellschaft sit alongside seven industrial partners including Amazon, the BMW Group, Daimler AG, IAV GmbH, Porsche AG, Robert Bosch GmbH, and ZF Friedrichshafen AG.

  The network also draws support from four foundations: the Christian Bürkert Foundation, the Gips-Schüle Foundation, the Vector Foundation, and the Carl Zeiss Foundation. That breadth of backing, from automotive giants to precision optics to public universities, gives Cyber Valley an unusual profile among European AI research consortia. The presence of both Stuttgart and Tübingen universities alongside MPI-IS means the network consolidates research capacity that already existed in the region rather than building it from scratch.

  The European Laboratory for Learning and Intelligent Systems, known as ELLIS, grew out of a related ambition. Founded in 2018, ELLIS set out to strengthen Europe's standing in global AI research, with MPI-IS as one of its anchors.