Event
Ph.D. Research Proposal: Srijal Shekhar Poojari
Thursday, April 30, 2026
12:00 p.m.
AVW 1146
ANNOUNCEMENT: Ph.D. Research Proposal Exam
Name: Srijal Shekhar Poojari
Committee:
Professor Derek Paley (Chair)
Professor Dinesh Manocha
Professor Kaiqing Zhang
Date/time: April 30, 2026 from 12 PM to 2 PM
Committee:
Professor Derek Paley (Chair)
Professor Dinesh Manocha
Professor Kaiqing Zhang
Date/time: April 30, 2026 from 12 PM to 2 PM
Location: A.V. Williams (AVW) Room 1146
Title: System Design and Integration of Autonomous Navigation and Active Sensing for Robotic Triage
Abstract:
This research presents the system design and implementation of a medical triage system using a legged robotic platform, specifically, the Boston Dynamics Spot robot, together with a sensing-equipped robotic arm. We detail the systems engineering of a specialized payload, focusing on the custom hardware, sensors, and subsystem integration that transform the base robot into a sensor platform for triage. To support field deployment, we also design the surrounding infrastructure, including robot control interfaces and robust communication networks.
During Mass Casualty Incidents, such as a natural disaster or a battlefield scenario, numerous victims with varying levels of injuries are dispersed across the scene of the event. Emergency medical responders are often overwhelmed in such incidents due to the number of victims, the complexity of the environment, or inherent hazards. This leads to inefficient care and the loss of life that could potentially be saved, motivating an increasing interest in robotic systems capable of supporting automated triage and casualty assessment.
Given the limited number of personnel available to remotely operate robotic assets during these large-scale events, it is essential to implement autonomy so that robots can operate independently while medics are en route or actively engaged with other casualties. Therefore, this research presents a framework for robot localization, path planning and navigation, enabling the robot to safely search the incident scene and reach casualties. We further address the transition of the robot from navigation to injury assessment through active sensing and visual servoing. In particular, we explore the use of a robotic manipulator for precise sensor placement, highlighting initial results in image-based visual servoing and outlining a research path toward fully autonomous casualty assessment.
This system has been extensively field tested on simulated mass casualty incidents as part of the DARPA Triage Challenge Systems Competition (2023-2026). Based on these real-world experiments, this proposal investigates the current strengths and limitations of the system, explores approaches in literature, and proposes specific hardware and software enhancements. Together, these research thrusts aim to provide a robotic system that assists medical personnel with the data required to prioritize life-saving interventions.
During Mass Casualty Incidents, such as a natural disaster or a battlefield scenario, numerous victims with varying levels of injuries are dispersed across the scene of the event. Emergency medical responders are often overwhelmed in such incidents due to the number of victims, the complexity of the environment, or inherent hazards. This leads to inefficient care and the loss of life that could potentially be saved, motivating an increasing interest in robotic systems capable of supporting automated triage and casualty assessment.
Given the limited number of personnel available to remotely operate robotic assets during these large-scale events, it is essential to implement autonomy so that robots can operate independently while medics are en route or actively engaged with other casualties. Therefore, this research presents a framework for robot localization, path planning and navigation, enabling the robot to safely search the incident scene and reach casualties. We further address the transition of the robot from navigation to injury assessment through active sensing and visual servoing. In particular, we explore the use of a robotic manipulator for precise sensor placement, highlighting initial results in image-based visual servoing and outlining a research path toward fully autonomous casualty assessment.
This system has been extensively field tested on simulated mass casualty incidents as part of the DARPA Triage Challenge Systems Competition (2023-2026). Based on these real-world experiments, this proposal investigates the current strengths and limitations of the system, explores approaches in literature, and proposes specific hardware and software enhancements. Together, these research thrusts aim to provide a robotic system that assists medical personnel with the data required to prioritize life-saving interventions.
