If we knew what it was we were doing, it would not be called research, would it? - Albert Einstein
Robotics and machine learning are crucial to saving lives by addressing the current and future challenges that the world and our planet are facing, such as major natural disasters (e.g., flood, hurricane, and earthquake), water and air pollution, global warming, etc. Swarm robotics is promising to provide effective and low-risk solutions to environmental monitoring and disaster response. Together with machine learning and optimization, they can minimize environmental footprints of (air and ground) vehicles in different domains such as transportation and logistics. As such, I decided to pursue my education in the field of robotics, machine learning, and optimization. I have been involved in four research projects, including two funded projects by DARPA and NSF on swarm robotics (coordination of a large number of collaborative robots) and two projects (one of them funded by UB) on machine learning and optimization (using machine learning models to reliably find an optimum design of expensive plants/simulations in a computationally efficient manner).
Selected Research Projects
Cognitive-Behavior Model to Predict Human Reaction to Swarm AI Non-Compliance
University at Buffalo, 01/2020 – Present
Swarm Tactics Design: Evolving Neural Architectures with Human Augmented Novelty for Complex Environments
University at Buffalo, 08/2019 – 08/2020
Machine Learning and Optimization Algorithms for Bio-inspired Design
University at Buffalo, 01/2017 – 06/2019
Investigating UAV Noise Impact on Human Hearing and Cognition
University at Buffalo, 06/2018 – 06/2019
Push Recovery for Humanoid Robots
University of Tehran, 09/2013 - 05/2015
University at Buffalo, 01/2020 – Present
- Extended swarm algorithms to allow human operators to modify them based on their needs or intention
- Created design of experiments for human subject studies
Swarm Tactics Design: Evolving Neural Architectures with Human Augmented Novelty for Complex Environments
University at Buffalo, 08/2019 – 08/2020
- Developed novel decentralized swarm algorithms for source localization by swarm robots via batch Bayesian search (see Bayes-Swarm) and scalable multi-robotic task allocation via graph theory (see BiG-MRTA)
- Investigated learning and imitation learning methods (learning by observing experts) to learn robust swarm tactics
- Led a group of four PhD students involved in project to prepare monthly reports for DARPA
- Kept track of deadlines and deliverables
Machine Learning and Optimization Algorithms for Bio-inspired Design
University at Buffalo, 01/2017 – 06/2019
- Developed a novel variable-fidelity optimization based on particle-based and batch Bayesian optimization
- Integrated a fully automated CFD framework, utilizing high performance computing resources using MPI, to find optimum riblet design for maximizing aerodynamic efficiency
- More
Investigating UAV Noise Impact on Human Hearing and Cognition
University at Buffalo, 06/2018 – 06/2019
- Designed a specialized experimental setup for testing impact of noise produced from small multi-rotor quadcopters on hearing and cognitive performance of humans
- Carried out design of experiments to conduct human subject study
- Determined significant features that impact hearing using statistical analysis
- More
Push Recovery for Humanoid Robots
University of Tehran, 09/2013 - 05/2015