What it does
The soft robotic gripper is designed to gently and securely grasp delicate items such as wine glasses by using soft silicone materials with smart sensor driven pneumatic system. Solving the problem of safely manipulating fragile objects in F&B industries.
Your inspiration
This inspiration for the project came from observing the limitations of traditional robotic grippers in handling delicate items. During my visit to a Bionic bar on a cruise, I observed an incident where a wine glass fell off the shelf, but the robotic system did not respond nor take any action to address it. I was motivated by the challenge of improving safety and adaptability in robotic manipulation. Using inspiration from nature such as octopus tentacles combining with the use of soft robotics, the idea grew to bridge the gap between rigid industrial robots and the need for careful, human like handling in any environment to practice safety.
How it works
The soft gripper is made from a mixture of specially chosen silicone materials, a mixture composition of Ecoflex 00-30 and DragonSkin 00-20 silicone moulded into a three fingered structure. These fingers were then inflated using a small air pump using positive pressure controlled by an Arduino microcontroller. A pressure sensor is attached to the fingers to continuously monitors the force applied by the gripper, allowing the system to automatically adjust the air pressure to maintain a secure, gentle hold on the objects. While also ensuring not too much pressure is applied such that it will break the fragile structure. If the pressure applied exceeds a certain threshold, a solenoid valve releases air to prevent over gripping and potential damage. This entire process is automated, ensuring the gripper adapts in real time to the object shape and fragility with no need for human intervention.
Design process
1. Concept and research - We reviewed existing soft gripper designs and actuation methods, identify the need for improve material selection, CAD drawing of the mould for the structure of the finger gripper. 2. Prototype 1 gripper using Ecoflex 00-30 silicone. Integrated flex sensors to experiment how much bent is needed to pick up the items. However, we found the Ecoflex 00-30 to be too weak for heavier objects and also flex sensors proved to be unsuitable as it has little potential for enhancement. 3. Prototype 2 using DragonSkin 00-20 for greater strength and durability. We also conducted structured weight lighting tests to see how much weight it can pick up, which confirms superior performance with fragile items. Furthermore, we replaced flex sensors with pressure sensor for more reliable feedback data and testing. 4. System integration - Developed Arduino based control logic for automated inflation and deflation, as well as automated pressure feedback control. 5. Iteration and testing - Refined the soft silicone gripper material by mixing a composition of Ecoflex and DragonSkin for better composition between flexibility and rigidity. Furthermore, improved reliability and repeatability of our system through multiple test cycles and real-world object handling.
How it is different
1. Unlike other soft grippers, our design directly compares and selects the optimal silicone material for both flexibility and strength, by mixing composition between a more solid silicone and a softer silicone material, 80% to 20% Dragonskin to Ecoflex composition. Resulting in better grip performance, durability and bent flexibility. 2. Real time feedback. This integration of pressure sensor and automated control logic from Arduino enables the gripper to adapt to different objects' pressure threshold, minimising the risk of breakage or slippage. It ensures just the right amount of pressure is applied to pick up the object and auto inflating and deflating mechanisms to adjust itself. 3. This system is also engineered for real world use. With a focus on easy maintenance, safe operation and consistent performance. We tested with real world wine glass and cups.
Future plans
1. We plan for more advance integrated sensors for more precise feedback and object detection. Furthermore, machine learning algorithm integrated to identify the object being picked up and adjust the gripper’s configuration, analysing the object’s size, shape, and estimated weight, it can determine the most effective gripping strategy tailored to each item. 2. Expand testing in real world environments such as food processing industries, healthcare or automated retails. 3. Refine the prototype for manufacturability and work together with companies to bring the gripper to market as a safe, reliable solution for delicate object manipulation.
Share this page on