The human hand is one of the most complex and essential systems in daily life, enabling dexterity and interaction. However, workers in industries such as manufacturing and construction are at high risk of hand injuries, while individuals who lose hand function face significant challenges. In the era of Industry 4.0, robotic hands are increasingly important in both industry and healthcare, yet their high cost remains a barrier due to expensive research and durable materials.

To address this, a low-cost robotic hand was developed in this project. The system is controlled by an ESP32 microcontroller, with an Arduino Nano handling the Voice Recognition Module V3. The hand performs ten common gestures through servo motors, while pressure sensors embedded in the thumb, index, and middle fingers provide tactile feedback to enhance grasp reliability.

Testing results show that the robotic hand can stably grasp cylindrical or spherical objects with diameters of 2–6 cm and weights of 100–200 g. Elastic objects, such as small bags, were easier to hold, while oversized or heavy items often slipped due to the limited closing range of the fingers. The hand was also able to perform ten predefined gestures in response to voice commands, although the offline recognition module reduced accuracy due to limited training capacity.

This project successfully achieved its objectives by building a voice-controlled robotic hand capable of mimicking basic human gestures and grasping everyday objects. Its modular and scalable design supports future improvements such as online voice recognition, Bluetooth or Wi-Fi control, or advanced EMG and AI-based gesture learning. Ultimately, this work contributes to affordable assistive technologies and supports the Sustainable Development Goals include SDG 3: Good Health and Well-being, and SDG 10: Reduced Inequalities).

Invention Technical Description

The Voice-Controlled Robotic Hand is developed as a low-cost solution to replicate essential human hand functions, focusing on gesture demonstration and basic object grasping. It addresses the challenge of expensive commercial robotic hands while providing potential applications in industrial automation, rehabilitation, and assistive devices for individuals with hand disabilities.

The hardware system integrates an ESP32 microcontroller as the main controller and an Arduino Nano that manages the Elechouse Voice Recognition Module V3. Recognised commands are transmitted to the ESP32, which then drives five servo motors to actuate the fingers. Force sensors embedded on the thumb, index, and middle fingers mimic tactile sensing, providing feedback to enhance grasp reliability and minimise slippage. A dedicated 5 V, 8 A power supply ensures stable operation of the servo motors and sensors. The outer casing and finger structures are produced using 3D printing with PLA filament to reduce manufacturing costs, while JST XH 2.54 mm connectors are used for circuit connections on a prototyping board to improve modularity and ease of maintenance.

The robotic hand operates through three integrated functions: control, power, and input/output. Voice commands recognised by the Nano are translated into servo movements by the ESP32. The power unit provides consistent energy to the motors, while the input and output system connects the tactile sensors and the servo motors, enabling interaction between the user’s command and the hand’s physical response.

Functionally, the robotic hand is able to perform ten predefined gestures, including numeric gestures such as ‘one’, ‘three’, ‘four’, and ‘five’, symbolic gestures like ‘okay’, ‘good’, and ‘finger gun’, and game gestures such as ‘rock’, ‘paper’, and ‘scissors’. In grasping tests, it reliably held objects with diameters between 2 to 6 cm and weights of 100 to 200 g. Items with elastic or compressible surfaces, such as a small ball, demonstrated greater stability when grasped, while oversized or heavier objects often slipped out due to the limited range of finger closure.

In conclusion, the Voice-Controlled Robotic Hand successfully meets its design objectives, offering a low-cost, modular platform capable of performing gestures and grasping everyday objects.

Demostration/ Presentation Video

Poster/ Broucher/ Invention Photo