Flex Sensor Gesture-Controlled Robotic Arm with Servo Feedback and Gripper Control using Arduino

The Flex Sensor Gesture-Controlled Robotic Arm with Servo Feedback and Gripper Control using Arduino lets a wearer steer a tabletop robotic arm by bending fingers inside a sensor glove, without a joystick, remote, or keyboard. Each glove finger carries a flex sensor whose electrical resistance rises as the finger bends, because bending stretches the sensor conductive layer and narrows its current path. An Arduino Uno reads all five sensors through voltage dividers, converts each reading into a servo angle, and drives five servo channels for base rotation, shoulder, elbow, wrist, and gripper motion. Servo feedback closes the loop: the controller compares each commanded angle with the fed-back joint position and corrects drift before the next move, so the gripper holds its pose instead of creeping. The result is a hands-free pick-and-place demonstrator that teaches sensor interfacing, analog conversion, servo control, and closed-loop feedback in one visible system.

The Flex Sensor Gesture-Controlled Robotic Arm with Servo Feedback and Gripper Control using Arduino works end to end from finger bend to arm motion as follows: bending a finger bends its flex sensor, the sensor resistance rises, and the voltage divider turns that resistance change into a changing analog voltage. An Arduino Uno reads that voltage, converts it to a digital value, maps it to a servo angle using the wearer calibration, and sends the angle to the matching servo, which rotates its joint to mirror the finger. Servo feedback then reports the joint position back, the controller corrects any drift, and the gripper follows the thumb: closing the thumb closes the gripper, and opening the thumb releases the held object. Calibration sits between reading and motion, so a straight finger and a bent finger map to opposite ends of the servo range, and smoothing plus a deadband stops sensor noise from shaking the arm between gestures.

1. The wearer bends a finger, bending its flex sensor and raising the sensor electrical resistance.
2. Each sensor sits in a voltage divider with a fixed resistor, so the resistance change becomes a changing analog voltage.
3. The Arduino Uno reads all five voltages on analog inputs using its 10-bit analog-to-digital converter.
4. Firmware maps each reading between the calibrated rest value and full-bend value to a servo angle from 0 degrees to 180 degrees, applying a moving average and deadband.
5. The Arduino sends the mapped angle to the matching servo channel: little finger to base, ring finger to shoulder, middle finger to elbow, index finger to wrist, and thumb to gripper.
6. Servo feedback is read back and compared with the commanded angle; the controller re-issues a correction when the error passes the deadband.
7. The gripper closes as the thumb bends and opens as the thumb straightens, holding its pose under feedback until the next gesture.

Builder note from the build: when the glove was first wired, the middle and ring sensor leads were swapped at the breadboard, so bending the middle finger moved the shoulder; labelling each lead by finger before powering up fixed it, and that labelling step is now part of the wiring checklist.

Tech stack:

• Arduino Uno
• Arduino IDE · C++ firmware
• Flex sensors · 5 sensor channels
• Servo motors · 5 servo channels
• Voltage dividers · 10 kΩ resistors
• Breadboard · jumper wiring
• External 5 V servo supply
• Serial Monitor readout

The Flex Sensor Gesture-Controlled Robotic Arm with Servo Feedback and Gripper Control using Arduino combines five-finger gesture input, five-channel servo drive, and closed-loop servo feedback into one hands-on system. Five flex sensors on a glove give one control channel per finger, so base, shoulder, elbow, wrist, and gripper can each be steered separately instead of moving together on a single switch. Servo feedback reports joint position back to the controller, per-wearer calibration adapts the glove to different hands, and smoothing plus a deadband keeps motion steady rather than jittery. A serial readout shows target and fed-back angles side by side, so a fault in one finger channel can be traced to one sensor, one divider, or one servo. Together, these capabilities let a student demonstrate sensing, signal conditioning, embedded firmware, and feedback control in a single pick-and-place prototype that can be tested joint by joint, without any wireless link, camera, or muscle-signal processing.

• [Five-finger gesture glove] Five flex sensors, one per finger, are sewn into a glove so each finger bend becomes a separate control signal for the arm.
• [Five-channel servo arm] Five servo channels drive base rotation, shoulder, elbow, wrist, and a two-finger gripper, giving four arm joints plus gripper control.
• [Servo position feedback] The controller compares commanded angles with fed-back joint position on the shoulder and gripper channels and corrects drift, while the remaining channels use commanded-angle tracking with smoothing.
• [Per-wearer calibration] Each sensor is calibrated at rest and at full bend for the wearer, because rest readings shift with glove fit and hand size.
• [Smooth, jitter-free motion] A moving average and a small deadband filter out sensor noise, so the arm moves steadily instead of twitching on small finger tremors.
• [Safe pick-and-place gripper] The two-finger gripper uses padded fingers and servo angle limits, and is intended only for small, light objects in a demonstration.
• [Serial monitor readout] Target angles and fed-back angles are printed to the Arduino Serial Monitor, so each control link can be checked during a demonstration.

Implemented vs optional: all seven features above are in the base build. Optional add-ons, not in the base build: a wireless glove link and a record-and-replay gesture sequence. Future scope, not built: force sensing in the gripper and a second arm for two-handed tasks.

• Working sensor glove with 5 flex sensors
• Tabletop robotic arm with 5 servo channels and two-finger gripper
• Arduino firmware with calibration, smoothing, deadband and servo feedback
• Circuit and wiring documentation with per-finger lead labelling
• Component list and setup instructions
• Project report PDF (background, sensor theory, methodology, results)
• PPT presentation for final review
• Demonstration guidance for pick-and-place tasks