Design and Implementation of an Autonomous Mobile Manipulator for Hospital Logistics
Primary Investigator (PI) Name
Razvan Voicu
Department
SPCEET – Robotics and Mechatronics Engineering
Abstract
A new wave of practical robotics is poised to lighten repetitive, time-critical workloads in clinical settings; one compelling use is a compact corridor courier that shuttles medications, samples, and supplies on demand. The problem is straightforward: staff time is routinely consumed by short-haul deliveries, which adds fatigue, slows care, and can elevate contamination risk, especially during peak hours and after shifts. This research proposes a feasibility study of a small indoor delivery robot that navigates a mapped hallway loop, accepts wireless dispatches, and docks reliably at designated stations. The approach seeks to combine ROS 2 (Nav2) for localization and routing with LiDAR/depth sensing for obstacle avoidance and a lightweight vision pipeline using fiducials and color cues for station identification; a simple top-mounted bin with a servo latch enables quick, touch-light handoffs without full manipulation. A LiDAR measurement test characterizes and validates the unit’s distance accuracy under various indoor conditions, ensuring dependable perception and mapping performance demonstrating its feasibily. Simulation in Gazebo precedes taped-corridor trials to tune parameters and reduce risk, while a Wi-Fi interface (MQTT/REST) triggers jobs and logs telemetry for reproducibility. Expected outcomes include end-to-end autonomous runs between two to three delivery points, consistent docking within practical handoff distances, smoother path tracking with fewer manual interventions, safer human-aware behavior in mixed traffic, and clearer baseline metrics for run time, route reliability, and recovery from common failure modes (missed tag, occlusion, lighting changes). The benefit is a tangible reduction in ad-hoc staff errands and a cleaner, reproducible template for short-haul logistics on edge compute. Future work will address dynamic re-planning in congested areas, secure payload handling with chain-of-custody, and multi-floor routing through elevator integration, supporting broader deployment across healthcare and laboratory environments.
Disciplines
Robotics
Design and Implementation of an Autonomous Mobile Manipulator for Hospital Logistics
A new wave of practical robotics is poised to lighten repetitive, time-critical workloads in clinical settings; one compelling use is a compact corridor courier that shuttles medications, samples, and supplies on demand. The problem is straightforward: staff time is routinely consumed by short-haul deliveries, which adds fatigue, slows care, and can elevate contamination risk, especially during peak hours and after shifts. This research proposes a feasibility study of a small indoor delivery robot that navigates a mapped hallway loop, accepts wireless dispatches, and docks reliably at designated stations. The approach seeks to combine ROS 2 (Nav2) for localization and routing with LiDAR/depth sensing for obstacle avoidance and a lightweight vision pipeline using fiducials and color cues for station identification; a simple top-mounted bin with a servo latch enables quick, touch-light handoffs without full manipulation. A LiDAR measurement test characterizes and validates the unit’s distance accuracy under various indoor conditions, ensuring dependable perception and mapping performance demonstrating its feasibily. Simulation in Gazebo precedes taped-corridor trials to tune parameters and reduce risk, while a Wi-Fi interface (MQTT/REST) triggers jobs and logs telemetry for reproducibility. Expected outcomes include end-to-end autonomous runs between two to three delivery points, consistent docking within practical handoff distances, smoother path tracking with fewer manual interventions, safer human-aware behavior in mixed traffic, and clearer baseline metrics for run time, route reliability, and recovery from common failure modes (missed tag, occlusion, lighting changes). The benefit is a tangible reduction in ad-hoc staff errands and a cleaner, reproducible template for short-haul logistics on edge compute. Future work will address dynamic re-planning in congested areas, secure payload handling with chain-of-custody, and multi-floor routing through elevator integration, supporting broader deployment across healthcare and laboratory environments.