Embedded Files
Our Project
Our Project
The Teleoperated Hand project was inspired by the many dangerous environments in which humans still perform tasks requiring dexterity beyond the capabilities of most current robots. While robots already handle some hazardous work, many jobs still depend on human hands for precise manipulation. Examples include confined low-oxygen spaces, high-voltage electrical systems, wind turbines and cell towers, chemical labs, mechanic shops, radioactive cleanup zones, and biohazard environments. In these situations, sending a human operator into danger creates significant risk, while remote robotic manipulation could greatly improve safety.
The core challenge is enabling robots to perform delicate, human-like tasks through effective telepresence. Operators need more than video feeds; they need responsive control and real-time sensory feedback such as touch, contact force, and resistance. By combining a robotic hand with haptic feedback, the system aims to give users a more natural sense of interacting with remote objects, allowing them to complete tasks that would otherwise require direct human presence.
Beyond safety, the concept also offers productivity and economic benefits. Skilled operators could control robots across multiple locations without travel, allowing faster response times and more efficient use of expertise. Companies could reduce costs related to training, protective equipment, insurance, downtime, and risk management. In the long term, data collected from teleoperated tasks could also help train increasingly autonomous robotic systems.
To keep the project focused and achievable, the prototype was scoped around lightweight objects, moderate grasp sizes, and reliable low-latency communication. The robotic hand emphasizes fine manipulation over heavy lifting, uses a practical multi-finger gripping design, and is powered through a wired supply during development. These assumptions allow the team to prioritize precision control, intuitive feedback, and system reliability, the key foundations of a useful teleoperated robotic platform.
System Bench Setup
REQUIREMENTS
Our Requirements are separated into functional and non-functional. Functional requirements are typically objective metrics of the robot system, and non-functional requirements are typically quantitative. All requirements have IDs that are referenced on the following spreadsheet. Verification and Validation tests will reference which requirements they address.
CONCEPT DESIGN
The concept combines three main subsystems: a wearable glove, a robotic hand mounted to an arm, and a central communication computer. As the user moves their hand, sensors in the glove capture finger motion, wrist orientation, and grip force. That data is processed and sent to the robotic hand, which mirrors the user’s movements. At the same time, sensors on the robot detect pressure, contact, and motion constraints, sending information back to the glove to create haptic feedback so the operator can feel interactions with objects. The system was designed for tabletop manipulation tasks such as grasping, lifting, placing objects, and reproducing hand gestures
A major focus of the concept design was balancing performance, safety, and usability. The team selected a centralized computing architecture where a computer runs ROS 2 and manages motion translation, communication, and logging, while local microcontrollers handle low-level actuator control. Safety systems were prioritized throughout the design, including emergency stop functions, command validation, fallback behavior in the event of communication loss, and limits on unsafe motions. The target control-loop latency was under 200 milliseconds to preserve a natural teleoperation experience.
Several trade studies guided the final concept. Electric motors were chosen for finger actuation because of their precision, repeatability, and compact integration compared with pneumatic or hydraulic systems. For user hand tracking, a hybrid approach was selected that combines glove sensors with vision-based sensing to improve reliability and reduce line-of-sight limitations. Fingertip pressure-sensing options, such as force-sensitive resistors and capacitive sensors, were planned for prototyping to determine the best balance of accuracy, durability, and cost.
Overall, the concept demonstrates a practical and scalable approach to teleoperated manipulation. By combining intuitive wearable controls, responsive robotic motion, and bidirectional haptic feedback, the system aims to create a more natural connection between human intent and robotic action.
TESTING VIDEOS
IMG_2530.movLoss of Tracking/Comms Test
IMG_2571.MOVReal Time Following Test
IMG_6038.MOVRobot Hand Mobility Test 1
Hand_Mobility_v2.MOVRobot Hand Mobility Test 2
IMG_5696.movServo Current Test
(Component Tests)
Glove _to_Hand_Control.movRobotic Hand Control Test
Arm_Realtime_v2.movRobotic Arm Control Test
Glove_Haptics_v1.movHaptic Feedback Test
References
The robot arm that we are using is a 5 DOF robot arm borrowed from the medical robotics class.
The wearable device is called Lucid Glove, it is an open source haptic feedback glove.
Page updated
Google Sites
Report abuse