Organizers

Amir Degani Amir Degani

Technion - Israel Institute of Technology
Howie Choset Howie Choset

Carnegie Mellon University
Justin Werfel Justin Werfel

Harvard University
Bhaskar Vundurthy Bhaskar Vundurthy

Carnegie Mellon University
Geordan Gutow Geordan Gutow

Carnegie Mellon University
Federico Oliva Federico Oliva

Technion - Israel Institute of Technology

Speakers

Karen Lee Bar-Sinai Karen Lee Bar-Sinai

Harvard Graduate School of Design

Dr. Karen Lee Bar-Sinai is an Assistant Professor of Materials and Design in Landscape Architecture and the Harvard University Graduate School of Design. She is a licensed architect, an urbanist, and holds a Ph.D. and postdoctoral training in robotic construction with found matter. Her GSD research group investigates the interaction between tools, materials, and the environment aiming to shift how and with what we build in the face of imminent material scarcity, environmental challenges, and climate change. Her research spans from small through territorial to planetary scales, all involving the modulation of matter in architectural or landscape construction. Current projects include Environmental Robotics – starting with ‘beaver-bots’ – simulating and developing beaver-inspired tools for restoring wetlands, with applications also in beaver-less sites like erosion-prone arid regions. In addition, she explores ways to deploy living materials – from fungi to root systems – as instruments for construction. She is also advancing Planetary Design Computation, testing the potential of targeted local landscape design to influence global climate-system dynamics. Karen Lee has lectured broadly on architecture, landscape architecture, and technology. She currently teaches core studios in landscape architecture and ‘Eco-Machina’ – a seminar on the emerging relationships between machines and landscapes.

Website
Spring Berman Spring Berman

Arizona State University

Spring Berman is an Associate Professor of Mechanical and Aerospace Engineering and Graduate Faculty in Computer Science, Electrical Engineering, and Exploration Systems Design at Arizona State University (ASU). She directs the Autonomous Collective Systems Laboratory and is an Associate Director of the Center for Human, Artificial Intelligence, and Robot Teaming (CHART) within the ASU Global Security Initiative. Prior to joining ASU in 2012, she was a postdoctoral researcher in Computer Science at Harvard University. She received the Ph.D. and M.S.E. degrees in Mechanical Engineering and Applied Mechanics from the University of Pennsylvania and the B.S.E. degree in Mechanical and Aerospace Engineering from Princeton University. She was a recipient of the ONR Young Investigator Award (2016) and the DARPA Young Faculty Award (2014). Her research focuses on the synthesis of scalable control strategies, including bio-inspired controllers, for robotic swarms and other types of distributed systems.

Website
Mark Yim Mark Yim

University of Pennsylvania
Sven Koenig Sven Koenig

University of California, Irvine

Sven is Chancellor's Professor and Bren Chair at the University of California, Irvine and a Fellow of AAAI, AAAS, ACM, and IEEE. Additional information about him can be found on his webpages

Website
Guillaume Sartoretti Guillaume Sartoretti

National University of Singapore

Guillaume Sartoretti joined the Mechanical Engineering Department at the National University of Singapore (NUS) as an Assistant Professor in 2019, where he founded the Multi-Agent Robotic Motion (MARMot) lab. Before that, he was a Postdoctoral Fellow in the Robotics Institute at Carnegie Mellon University (USA), where he worked with Prof. Howie Choset. He received his Ph.D. in robotics from EPFL (Switzerland) in 2016 for his dissertation on "Control of Agent Swarms in Random Environments," under the supervision of Prof. Max-Olivier Hongler. His passion and research lie in understanding and eliciting emergent coordination/cooperation in large multi-agent systems, by identifying what information and mechanisms can help agents reason about their individual role/contribution to each other and to the team. Guillaume was a Manufacturing Futures Initiative (MFI) postdoctoral fellow at CMU in 2018-2019, was awarded an Amazon Research Awards in 2022, as well as an Outstanding Early Career Award from NUS' College of Design and Engineering in 2023.

Website
Karola Dierichs Karola Dierichs

Max Planck Institute of Colloids and Interfaces, Department of Biomaterials

Karola Dierichs holds the Material and Code professorship in the Cluster of Excellence Matters of Activity. In addition, she is a researcher at the Max Planck Institute of Colloids and Interfaces. Her expertise lies in the fields of materials design and minimal machines for architectural construction. Here, the main goal is to establish architecture as sourced from and embedded in a given environment. For this, methods of science and art are integrated to establish a novel paradigm of fundamental research. Previous affiliations include the Institute for Computational Design and Construction at the University of Stuttgart, where she has conducted research in the field of designed granular materials for architecture.

Website

Schedule

Room 315

Morning

08:30 - 09:30 Welcome & Opening remarks
09:30 - 10:30 Keynote Speakers
Collective Construction in Landscape Architecture

Abstract

This presentation explores the emergence of collective robotic tools in landscape architecture—such as autonomous earthwork systems, extraterrestrial construction tools for in-situ resource utilization (ISRU), and swarm robotics for terrain and erosion management. We argue that these technologies represent a shift from top-down design toward situated, adaptive construction that dynamically responds to environmental conditions. Drawing inspiration from natural systems—particularly beavers as ecosystem engineers—we reframe autonomy as a distributed, ecological process. Beavers collaboratively reshape landscapes through simple, decentralized actions, resulting in complex constructs, reconfigured hydrological systems, and enriched ecologies. We will present early research into beaver-inspired tools and envision a future autonomous system composed of responsive agents working with, rather than against, ecological systems and environmental flows. By learning from natural collectives, we propose a reframing of environmental robotics as a co-evolutionary practice—integrating robotics, ecology, and design. This nascent field uses robotic tools to facilitate collaborative environmental processes and cultivate new synergies with the environment.
Karen Lee Bar-Sinai
Designing Multi-Agent Matter

Abstract

Granular materials such as sand are modular systems: a large number of units—for example sand grains—are in loose contact with each other. This loose interaction allows for the formation of solid, liquid and gaseous states in granular materials. The characteristics of a granular material are defined by the materiality and geometry of its component units—the particles. If these particles are designed, entirely new characteristics can be programmed into a granular material. Departing from this notion of a »designed granular material« the talk will show how granular materials can become multi-agent matter. Moving from self-interlocking particles for architecture-scale construction to autonomously entangling ones—we will pose the question how matter itself can become a robotic system.
Karola Dierichs
10:30 - 10:45 Coffee Break
10:45 - 11:45 Keynote Speakers
Docking Mechanisms for Lunar Surface Technology

Abstract

This talk will deal with the progress in the NASA TRUSSES project for docking mechanisms for robots to attach to each other in the context of Lunar regolith interaction. This project explores methods for teams of robots to jointly overcome environmental hazards on the Moon by attaching to each other to form larger and more stable, maneuverable structures. In the process, we can explore ground interactions required for building regolith structures and site preparation on the moon.
Mark Yim
Planning with the TERMES Robots

Abstract

Cooperative multi-robot planning is an understudied but important area of AI planning due to the increasing importance of multi-robot systems. The Harvard TERMES robots can build 3D structures by picking up individual blocks, carrying them around, putting them down, and climbing them. Planning for single robots is already difficult due to the large number of blocks and long plans. Planning for multiple robots is even more difficult since it needs to reason about how to achieve a high degree of parallelism without agents obstructing each other, even though many robots operate together in tight spaces. In this talk, I describe previous research on a first (domain-dependent, centralized, and non-optimal) multi-agent planning method for this domain that compared favorably to off-the-shelf planning technologies.


Sven Koenig
11:45 - 12:30 Lightning talks

Afternoon

12:30 - 13:30 Lunch Break
13:30 - 14:00 Poster session
14:00 - 15:00 Keynote Speakers
Decentralized Strategies for Multi-Robot Object Transport and Collision-Free Navigation

Abstract

Robot collectives that perform construction tasks must be able to safely navigate around changing arrangements of materials and cooperatively transport objects that are too large or heavy to be moved by a single robot. To achieve collision-free robot navigation in environments with unknown obstacles, we have designed nonlinear model predictive control methods, integrating barrier functions learned from the robot’s sensor measurements, as well as virtual potential-based controllers. We have also developed decentralized robot controllers for cooperative transport that, unlike prior approaches, do not rely on inter-robot communication or prior information about the object, environment, or robot transport team; each robot only requires the target object location or velocity and measurements of its own state. Some of these controllers were designed to reproduce observed features of group food retrieval in desert ants, with the aim of producing similarly robust transport behaviors. These navigation and control strategies are demonstrated in numerical and physics-based simulations and in experiments with small mobile ground robots.


Spring Berman
Towards Learned Cooperation at Scale in Robotic Multi-Agent Systems

Abstract

With the recent advances in sensing, actuation, computation, and communication, the deployment of large numbers of robots is becoming a promising avenue to enable or speed up complex tasks in areas such as manufacturing, last-mile delivery, search-and-rescue, or autonomous inspection. My group strives to push the boundaries of multi-agent scalability by understanding and eliciting emergent coordination/cooperation in multi-robot systems as well as in articulated robots (where agents are individual joints). Our work mainly relies on distributed (multi-agent) reinforcement learning, where we focus on endowing agents with novel information and mechanisms that can help them align their decentralized policies towards team-level cooperation. In this talk, I will first summarize my early work in independent learning, before briefly discussing my group's recent advances in convention, communication, and context-based learning. Throughout this journey, I will highlight the key challenges surrounding learning representations, policy space exploration, and scalability of the learned policies, and outline some of the open avenues for research in this exciting area of robotics.
Guillaume Sartoretti
15:00 - 15:15 Coffee Break
15:15 - 16:30 World Café
16:30 - 17:00 Open discussion
17:00 - 17:30 Feedback and Closing Remarks