Swarm Tactics Learning and Human-Swarm Interaction

Read More

Talent Metrics Sequential Co-design

Read More

Multi-Robot Task Allocation using Graph Learning

Read More

Stochasticity and Generalization of Autonomous Operation

Read More

Swarm Robotics

Robotic systems, often working collaboratively as a team of multiple autonomous agents, are becoming valuable players in different real-world applications; these… Read More

Neuroevolution

In decision-support applications for autonomous systems, the architecture or topology of Artificial Neural Networks (ANN) is often user-prescribed, thereby leading to sub-optimal prediction models…

Read More

Transitioning UAVs

Multi-ability unmanned aerial vehicles have grown attention in the last decade. They also provide the advantage of vertical take-off and landing and forward flight…

Read More

Homepage

LATEST NEWS & EVENTS

    • Congrats to undergrad researcher, Eric Butcher (CSE sophomore) for being selected to the ONR Naval Research Enterprise Internship Program (NREIP), Summer 2024.
    • Congrats to Steve Paul, for successfully defending his Doctoral Dissertation on “Higher-Order Graph Reinforcement Learning for Multi-Agent Systems and Physical Networks“, Jan 3, 2024.
    • Dr. Chowdhury has been inducted by AIAA into the class 2024 AIAA Associate Fellows
    • New funding award — ONR, Science of AI, GRAPPLE: Generalizable & Robust Activity Planning enabled by Physics based Abstraction, Simulation & Learning environmentsPI: Souma Chowdhury, Period: Nov 2023 – Apr 2027.
    • Dr. Souma Chowdhury has been selected to join the Executive Committee of ASME DAC, in August 2023. The DAC Executive Committee is a rotating group of five experts in Design Automation that organizes the DA conference (DAC) under ASME IDETC.
    • New supplement funding — NSF CAREER REU Supplement (two): To support swarm robotics research of two undergraduate students (from MAE and CSE) as part of the following NSF project: ““Automated Design of Decentralized Robust and Explainable Swarm Systems (ADDRESS)“. July 2023 – July 2024.”
    • Congrats! to Chris Kakis (UG junior) for receiving the $2,500 Zimmer scholarship from UB for his ongoing experimental research work on tether net systems in ADAMS Lab.
    • New funding award — NASA, R&TD – Innovative Spontaneous Concepts (ISC), Physics-informed Machine Learning (PIML) towards a Real-time and Interpretable Thermal Simulator, Lead: NASA JPL, Sub-Contract: University at Buffalo (PI: Souma Chowdhury), Period: Apr 2023 – Aug 2023.
    • Prajit (PhD student) and Henry (UG senior) performed an exciting demonstration of AI-based Air-Traffic Control of four drones — for over 70 attendees in AIAA Region 1 Student Conference, Apr 1, 2023.
    • Congrats! to Nathan Maurer for graduating with his MS Thesis in Robotics, Jan 2023
    • Congrats! to Jhoel Witter for graduating with his MS Thesis in Robotics, Jan 2023
    • Dr. Chowdhury receives membership elevation to IEEE Senior Member, Sep 27, 2022.
    • Dr. Chowdhury starting on his 1 month sabbatical visit at NASA Jet Propulsion Lab (JPL), Maritime and Multi-Agent Autonomy group, Sep 26, 2022.

To conceive, analyze and design complex systems, we investigate new approaches that are founded on a fundamental notion ofadaptation”. Adaptation is realized by bringing together nature inspired principles of computation, rigorous engineering design methods, and machine learning.

Open Positions in ADAMS Lab (2022-2023):

ADAMS Lab is also looking for exceptional Masters and Undergraduate students who have the motivation to excel in the below areas of scholarly research. A successful student in these areas must have a solid background in mathematics, engineering, and computer programming. Good communication skills and previous research experience are a plus. For further information, click here, or contact Dr. Chowdhury: soumacho@buffalo.edu .

Fundamental areas of research in the ADAMS lab include:

  • Swarm Systems: swarm intelligence for distributed search, task allocation, optimization, and decentralized cyber-physical systems.
  • Evolutionary–Neural Algorithms: concurrent design of the morphological and intelligence architecture of autonomous systems; evolution of neural network topologies.
  • Physics-cognizant machine learning: physics-aware hybrid machine learning architectures for dynamic systems
  • Metamodeling and Multi-fidelity Optimization: automated selection and adaptive refinement of metamodels; multi-objective and mixed-integer optimization.

Major areas of application include:

  • Unmanned Aerial Vehicles (UAVs)
  • Swarm Robotics & Multi-Robotics
  • Metamaterial Systems Design & Morphological Computing Bio-inspired Flow Modulation 
  • Physics Infused Machine Learning
  • Neuro-evolution and Graph Learning
  • Cyber Physical Systems (e.g., real-world networks)
  • Urban Air Mobility aka Flying Cars

Most Recent Publications

Urban Air Mobility Vertiport Management

Urban Air Mobility (UAM) promises a new dimension to decongested, safe, and fast travel in urban and suburban hubs. These UAM aircraft are conceived to operate from small airports called vertiports each comprising multiple take-off/landing and battery-recharging spots. Since they might be situated in dense urban areas and need to handle many aircraft landings and take-offs each hour, managing this schedule in real-time becomes challenging for a traditional air-traffic controller but instead calls for an automated solution. We are working on a novel approach to this problem of Urban Air Mobility - Vertiport Schedule Management (UAM-VSM), which leverages graph reinforcement learning to generate decision-support policies. Here the designated physical spots within the vertiport's airspace and the vehicles being managed are… Read More

Neuro-evolution

Adaptive Genomic Evolution of Neural-Network Topologies (AGENT) Topology and weight evolving artificial neural network (TWEANN) algorithms optimize the structure and weights of artificial neural networks (ANNs) simultaneously. The resulting networks are typically used as policy models for solving control and reinforcement learning (RL) type problems. This paper presents a neuroevolution algorithm that aims to address the typical stagnation and sluggish convergence issuespresent in other neuroevolution algorithms. These issues are often caused by inadequacies in population diversity preservation, exploration/exploitation balance, and search flexibility. This new algorithm, called the Adaptive Genomic Evolution of Neural Network Topologies (AGENT), builds on the neuroevolution ofaugmenting topologies (NEAT) concept. Novel mechanisms for adapting the selection and mutation operations are proposed to favorably control population diversity and… Read More

Graph Learning Based solution for MRTA

Author: Steve Paul Efficient multi-robot task allocation (MRTA) is fundamental to various time-sensitive applications such as disaster response, warehouse operations, and construction. MRTA involves using a team of robots to perform tasks that are spatially distributed, and present time deadlines and different workloads that may require multiple visits by robots to complete the task. https://www.youtube.com/watch?v=X-M12l3M43s Here, we proposed encoder-decoder graph neural network involving Capsule networks and multi-head attention mechanism, and innovatively add topological descriptors (TD) as new features to improve transferability to unseen problems of similar and larger size. Persistent homology is used to derive the TD, and proximal policy optimization is used to train our TD-augmented graph neural network. The resulting policy model compares favorably to state-of-the-art non-learning baselines… Read More