AI based Robotic Coin Counting and Sorting System (2022)
Indian Patent No. 567408, dated 01 January 2023; SASTRA-TRR Fund (Rs 2 Lakh)
- Manual coin counting in banks and temples across India is slow, error-prone, and labor-intensive, especially when handling large volumes.
- The system automates the entire process using a high-quality microscopic camera, motorized conveyor, and a gantry robot with a suction gripper.
- The microscopic camera captures high-resolution images of each coin, the pre-trained AI model identifies the denomination and updates the count in real time, the conveyor moves the coin to the pickup zone, and the gantry robot uses its suction gripper to pick and deposit the coin into the designated bin.
- This automated setup drastically reduces manual effort, improves audit accuracy, and ensures fast, consistent, and reliable sorting, making it ideal for temples, banks, and institutions handling large volumes of coins.
- The project was supported by SASTRA-TRR grant amounting to Rs 2 Lakh.
- Successfully demonstrated at the Sabarimala Sree Ayyappa Temple, Kerala.
Design and Development of Novel Self-Transfer Devices for Wheel Chair Users (2012-17)
Indian Patent No. 371329, dated 07 July 2021; Funded by Department of Science & Technology,
Govt. of India (Rs 18 Lakh)
- Wheelchair users face great difficulty in transferring themselves from one surface to another, for example from wheelchair to a toilet commode. In such cases, mostly a caregiver’s assistance may be required, but it affects one’s dignity.
- A robotic self-transfer device and a manual self-transfer device aimed at offering privacy and independence to lower limb disabled in performing daily activities have been developed.
- The device- robotic or manual, attached to a wheelchair is useful in transferring a user from a wheelchair to a toilet commode or any another surface following simple and natural transfer procedure without the need of any caregiver.
- The user can achieve transfer by operating joysticks in the case of manual version and makes use of the available upper body strength of the user for transfer in the manual version.
- While the robotic version employs two linear actuators and a motor to accomplish the transfer, the manual version employs lockable gas springs and handwheel with epicyclic gear train to achieve transfer.
- Trials were carried out to test the overall acceptability of the devices by involving potential beneficiaries and devices could successfully transfer the participants from a wheelchair to a chair with less effort in less than a minute.
- The manual version is first of its kind in the world and it was also found that the robotic version is superior to the other existing transfer systems in terms of comfort and operation.
- The manual version is the World's first mechanized self-transfer device. The manual one and the powered one, both are India's first self-transfer devices for wheelchair users.
Conceptual design of the transfer process
Manual & Powered Self-Transfer devices
Transfer of a person from wheelchair to a chair using the powered self-transfer device
Demonstration of World's first manual-mechanized self-transfer device aired in Kalvipaarvai programme in DD Podhigai (June 2015)
Developing an 8 DoF Humanoid- EKINBOT and achieving Human Robot Interaction (2011-12)
Locomotion of a Humanoid robot with lesser number of actuators is a choice of interest, as it leads to energy efficient design. This work emphasis on design of a humanoid robot with minimal actuation, minimal control and minimum development cost. The work covered, design considerations and methods used for balancing and walking gait generation of a Humanoid Robot with 8 Degrees of Freedom. The robot finds a place in between the most sophisticated, complex humanoid robots and simple, miniaturized humanoids. Hand motion tracking and Gesture recognition are fundamental technologies for Human-Computer interaction. Same technology has been adapted for Human- Robot Interaction. For which a mechanism analogous to ‘Data Glove’ has been used. The main objective of the project was to design a simplified Humanoid Robot with lesser number of actuators, make it walk successfully, perform certain common robot applications like obstacle detection and finally the Human Robot Interaction.
