About

Taskin Padir is a Professor in the Electrical and Computer Engineering Department at Northeastern University and also serves as an Amazon Scholar. He holds a PhD and MS degrees from Purdue University, earned in 2004 and 1997 respectively, and a BS from Middle East Technical University in 1993. He was the Founding Director of the Institute for Experiential Robotics at Northeastern University. His research focuses on shared autonomy, human-in-the-loop robotics, humanoids, collaborative robotics, embodied artificial intelligence, and human-robot teaming at the extremes. Padir has led teams in significant robotics competitions such as the DARPA Robotics Challenge and NASA Sample Return Challenge. His work has been funded by prominent agencies including NSF, DARPA, NASA, DoD, DoE, ONR, USACE, and industry partners like Amazon Robotics, Verizon, MathWorks, and Intel. He has received multiple awards, including the 2024 Faculty Research Team Award, the Impact Award in 2023, and the Kalenian Award for Entrepreneurial Spirit, among others. Padir has also held leadership roles such as serving on the Executive Committee of the Robotics and Remote Systems Division of the American Nuclear Society and as an editor for the International Conference in Robotics and Automation. His research laboratory, the Robotics and Intelligent Vehicles Research Laboratory (RIVeR Lab), advances autonomous robots and intelligent vehicles, focusing on design, analysis, and control of various robotic systems aligned with real-world problems.

Research topics

• Artificial Intelligence
• Computer Science
• Computer vision
• Machine Learning
• Human–computer interaction
• Political Science
• Engineering
• Mathematics
• Psychology
• Multimedia
• Medicine
• Operating system
• Data science
• Systems engineering
• Remote sensing
• Gerontology
• Geology
• Simulation

Selected publications

• “Meet My Sidekick!”: Effects of Separate Identities and Control of a Single Robot in HRI

  2026-03-10

  articleOpen access

  The presentation of a robot's capability and identity directly influences a human collaborator's perception and implicit trust in the robot. Unlike humans, a physical robot can simultaneously present different identities and have them reside and control different parts of the robot. This paper presents a novel study that investigates how users perceive a robot where different robot control domains (head and gripper) are presented as independent robots. We conducted a mixed design study where participants experienced one of three presentations: a single robot, two agents with shared full control (co-embodiment), or two agents with split control across robot control domains (split-embodiment). Participants underwent three distinct tasks -- a mundane data entry task where the robot provides motivational support, an individual sorting task with isolated robot failures, and a collaborative arrangement task where the robot causes a failure that directly affects the human participant. Participants perceived the robot as residing in the different control domains and were able to associate robot failure with different identities. This work signals how future robots can leverage different embodiment configurations to obtain the benefit of multiple robots within a single body.

  PublisherOA PDFDOI

• Gesture-Based Human-in-the-Loop Control of Space Exploration Vehicle Convoys

  2026-03-07

  articleSenior author

  The deployment of multi-robot systems for planetary surface operations is essential as space exploration initiatives grow in scale and complexity. Coordinated convoys of rovers hold promise for a range of critical missions, including scientific exploration, material transport, infrastructure construction, and emergency response. However, enabling robust inter-rover communication and dynamic coordination in uncertain, unstructured environments presents technical and human factors challenges. This research is aimed at the development of a gesture-based control architecture tailored for human-in-the-loop space exploration vehicle convoys. The system demonstrates a framework in which a lead vehicle is driven by the human operator using a conventional steering wheel, while a trailing rover receives high-level commands via hand gestures detected through a wearable technology. This approach leverages intuitive interactions to minimize cognitive load while maintaining operational efficiency. When gesture input is not actively provided, the semi-autonomous rover defaults to a reactive mode, synchronizing its speed and trajectory with the lead vehicle to preserve formation and maintain convoy integrity. A Microsoft HoloLens 2 Heads-Up Display (HUD) provides contextual information to the operator via augmented reality (AR) visualizations. A real-time cognitive workload estimation module integrates cardiovascular physiological signals-specifically heart rate and heart rate variability-into a fuzzy logic framework for shared control in convoy operations. This enables real-time adaptation of vehicle control authority based on the driver's cognitive state. Effectiveness of this novel interface design has been validated in simulation and hardware utilizing small scale autonomous vehicles.

  PublisherDOI

• "Meet My Sidekick!": Effects of Separate Identities and Control of a Single Robot in HRI

  arXiv (Cornell University) · 2026-02-07

  preprintOpen access

  The presentation of a robot's capability and identity directly influences a human collaborator's perception and implicit trust in the robot. Unlike humans, a physical robot can simultaneously present different identities and have them reside and control different parts of the robot. This paper presents a novel study that investigates how users perceive a robot where different robot control domains (head and gripper) are presented as independent robots. We conducted a mixed design study where participants experienced one of three presentations: a single robot, two agents with shared full control (co-embodiment), or two agents with split control across robot control domains (split-embodiment). Participants underwent three distinct tasks -- a mundane data entry task where the robot provides motivational support, an individual sorting task with isolated robot failures, and a collaborative arrangement task where the robot causes a failure that directly affects the human participant. Participants perceived the robot as residing in the different control domains and were able to associate robot failure with different identities. This work signals how future robots can leverage different embodiment configurations to obtain the benefit of multiple robots within a single body.

  PublisherDOI

• Chance-Constrained Convex MPC for Robust Quadruped Locomotion Under Parametric and Additive Uncertainties

  IEEE Robotics and Automation Letters · 2025-07-07 · 3 citations

  articleSenior author

  Recent advances in quadrupedal locomotion have focused on improving stability and performance across diverse environments. However, existing methods often lack adequate safety analysis and struggle to adapt to varying payloads and complex terrains, typically requiring extensive tuning. To overcome these challenges, we propose a Chance-Constrained Model Predictive Control (CCMPC) framework that explicitly models payload and terrain variability as distributions of parametric and additive disturbances within the single rigid body dynamics model. Our approach ensures safe and consistent performance under uncertain dynamics by expressing the model's friction cone constraints, which define the feasible set of ground reaction forces, as chance constraints. Moreover, we solve the resulting stochastic control problem using a computationally efficient quadratic programming formulation. Extensive Monte Carlo simulations of quadrupedal locomotion across varying payloads and complex terrains demonstrate that CCMPC significantly outperforms two competitive benchmarks: Linear MPC and MPC with hand-tuned safety margins to maintain stability, reduce foot slippage, and track the center of mass. Hardware experiments on the Unitree Go1 robot show successful locomotion across various indoor and outdoor terrains with unknown loads exceeding 50% of the robot's body weight, despite no additional parameter tuning. A video of the results and accompanying code can be found at: <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://cc-mpc.github.io/</uri>.

  PublisherDOI

• Spectral Signature Mapping from RGB Imagery for Terrain-Aware Navigation

  2025-12-08

  articleSenior author

  Successful navigation in outdoor environments requires accurate prediction of the physical interactions between the robot and the terrain. Many prior methods rely on geometric or semantic labels to classify traversable surfaces. However, such labels cannot distinguish visually similar surfaces that differ in material properties. Spectral sensors enable inference of material composition from surface reflectance measured across multiple wavelength bands. Although spectral sensing is gaining traction in robotics, widespread deployment remains constrained by the need for custom hardware integration, high sensor costs, and compute-intensive processing pipelines. In this paper, we present the RGB Image to Spectral Signature Neural Network (RS-Net), a deep neural network designed to bridge the gap between the accessibility of RGB sensing and the rich material information provided by spectral data. RS-Net predicts spectral signatures from RGB patches, which we map to terrain labels and friction coefficients. The resulting terrain classifications are integrated into a sampling-based motion planner for a wheeled robot operating in outdoor environments. Likewise, the friction estimates are incorporated into a contact-force-based MPC for a quadruped robot navigating slippery surfaces. Overall, our framework learns the task-relevant physical properties offline during training and thereafter relies solely on RGB sensing at run time. The code is available at https://github.com/prajapatisarvesh/RS-Net.

  PublisherDOI

• Assessing the Impact of a Passive Exoskeleton on Firefighter Performance and Physiological Response

  2025-08-25

  articleSenior author

  Firefighters operate in hazardous environments with limited ergonomic support, often leading to significant physical strain. While robotics research has explored drones and quadrupeds for firefighting assistance, exoskeletons remain underutilized. This study evaluates the effects of the BackX passive exoskeleton during firefighter search and rescue tasks. Five professional firefighters performed rescue and equipment carry tasks with and without the exoskeleton, while physiological metrics were recorded using the COSMED K5 metabolic analyzer. Results showed a reduction in cardiovascular strain and anaerobic demand when using the exoskeleton; however, energy expenditure increased, likely due to restricted movement and inefficiencies. Post-task surveys indicated reduced perceived exertion and fatigue. These findings suggest that passive exoskeletons may alleviate physical demands but require further development to improve energy efficiency and usability in dynamic emergency scenarios. Continued research is necessary to optimize exoskeleton design for fire service applications and to assess long-term operational benefits.

  PublisherDOI

• Spectral Signature Mapping from RGB Imagery for Terrain-Aware Navigation

  ArXiv.org · 2025-09-23

  preprintOpen accessSenior author

  Successful navigation in outdoor environments requires accurate prediction of the physical interactions between the robot and the terrain. Many prior methods rely on geometric or semantic labels to classify traversable surfaces. However, such labels cannot distinguish visually similar surfaces that differ in material properties. Spectral sensors enable inference of material composition from surface reflectance measured across multiple wavelength bands. Although spectral sensing is gaining traction in robotics, widespread deployment remains constrained by the need for custom hardware integration, high sensor costs, and compute-intensive processing pipelines. In this paper, we present the RGB Image to Spectral Signature Neural Network (RS-Net), a deep neural network designed to bridge the gap between the accessibility of RGB sensing and the rich material information provided by spectral data. RS-Net predicts spectral signatures from RGB patches, which we map to terrain labels and friction coefficients. The resulting terrain classifications are integrated into a sampling-based motion planner for a wheeled robot operating in outdoor environments. Likewise, the friction estimates are incorporated into a contact-force-based MPC for a quadruped robot navigating slippery surfaces. Overall, our framework learns the task-relevant physical properties offline during training and thereafter relies solely on RGB sensing at run time.

  PublisherOA PDFDOI

• SCANS: A Soft Gripper With Curvature and Spectroscopy Sensors for In-Hand Material Differentiation

  IEEE Robotics and Automation Letters · 2025-10-13 · 1 citations

  article

  We introduce the soft curvature and spectroscopy (SCANS) system: a versatile, electronics-free, fluidically actuated soft manipulator capable of assessing the spectral properties of objects either in hand or through pre-touch caging. This platform offers a wider spectral sensing capability than previous soft robotic counterparts. We perform a material analysis to explore optimal soft substrates for spectral sensing, and evaluate both pre-touch and in-hand performance. Experiments demonstrate explainable, statistical separation across diverse object classes and sizes (metal, wood, plastic, organic, paper, foam), with large spectral angle differences between items. Through linear discriminant analysis, we show that sensitivity in the near-infrared wavelengths is critical to distinguishing visually similar objects. These capabilities advance the potential of optics as a multi-functional sensory modality for soft robots. The complete parts list, assembly guidelines, and processing code for the SCANS gripper are accessible at: <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://parses-lab.github.io/scans/</uri>.

  PublisherDOI

• robot_collision_checking: A Lightweight ROS 2 Interface to FCL (Flexible Collision Library)

  The Journal of Open Source Software · 2025-01-27 · 1 citations

  articleOpen accessSenior author

  This paper presents robot_collision_checking, a C++ library that provides a Robot Operating System (ROS) (Quigley et al., 2009) interface to the Flexible Collision Library (FCL) (Pan et al., 2012) for typical robotics applications.FCL is an open-source C++ library that provides efficient collision detection and distance computation for 3D environments.While these capabilities are crucial in robotics to ensure safety and enable effective motion planning, FCL is not readily available for many robot architectures built atop ROS.Given that the robotics community widely relies on ROS as the standard for software development, it would greatly benefit from a lightweight ROS interface to FCL.The robot_collision_checking package fulfils this demand by exposing FCL functionality to ROS message types, thereby allowing robotics researchers and practitioners that rely on ROS to easily access the collision and distance checking features of FCL.

  PublisherOA PDFDOI

• VOTE: Vision-Language-Action Optimization with Trajectory Ensemble Voting

  ArXiv.org · 2025-07-07

  preprintOpen access

  Recent large-scale Vision Language Action (VLA) models have shown superior performance in robotic manipulation tasks guided by natural language. However, current VLA models suffer from two drawbacks: (i) generation of massive tokens leading to high inference latency and increased training cost, and (ii) insufficient utilization of generated actions resulting in potential performance loss. To address these issues, we develop a training framework to finetune VLA models for generating significantly fewer action tokens with high parallelism, effectively reducing inference latency and training cost. Furthermore, we introduce an inference optimization technique with a novel voting-based ensemble strategy to combine current and previous action predictions, improving the utilization of generated actions and overall performance. Our results demonstrate that we achieve superior performance compared with state-of-the-art VLA models, achieving significantly higher success rates and 39$\times$ faster inference than OpenVLA with 46 Hz throughput on edge platforms, demonstrating practical deployability. The code is available at https://github.com/LukeLIN-web/VOTE.

  PublisherOA PDFDOI

Recent grants

• FW-HTF-RL: Co-worker Robots to Impact Seafood Processing (CRISP): Designs, Tools and Methods for Enhanced Worker Experience

  NSF · $2.5M · 2019–2024

• Enhancing Disabilities Engineering Research and Education Through Robotics Capstone Projects

  NSF · $87k · 2016–2018

• PFI:AIR-TT: LoCATER:Localization and Accountability Technology for Emergency Responders

  NSF · $200k · 2017–2018

• RAPID: Realization of a Medical Cyber-Physical System to Enhance Safety of Ebola Workers

  NSF · $200k · 2014–2016

• RAPID: Realization of a Medical Cyber-Physical System to Enhance Safety of Ebola Workers

  NSF · $54k · 2016–2016

Frequent coauthors

• Deniz Erdoğmuş

  Northeastern University

  24 shared

• Nathaniel Hanson

  Massachusetts Institute of Technology

  23 shared

• Philip Long

  23 shared

• Velin Dimitrov

  Toyota Research Institute

  20 shared

• Michael A. Gennert

  Worcester Polytechnic Institute

  15 shared

• Dmitry A. Sinyukov

  14 shared

• Aykut Özgün Önol

  14 shared

• Mark Zolotas

  Northeastern University

  12 shared

Labs

• Robotics and Intelligent Vehicles Research LaboratoryPI

Education

• PHD, Electrical and Computer Engineering

  Purdue University

  2004

Awards & honors

• 2024 Faculty Research Team Award
• 2023 Impact Award
• 2022 Amazon Scholar
• 2022 Faculty Research Team Award
• 2020 Faculty Research Team Award