Resumen de Design and evaluation of variable-stiffness robotic grippers
Daniel Cardin Catalan
In today's retailing industry, innovation and novelty are at the forefront of productivity and improvement. In most store warehouses, objects differ in terms of shape and weight. During the automation process for the selection of individual objects, reliable grasping is required, due to the variability of object shapes and positions. If one gripper can be used for all of the objects in the warehouse, this would increase productivity levels and profits considerably.
The main objective of this thesis dissertation is to create, analyse and validate a specific gripper design. This gripper should be able to adapt to unknown objects, grasping them firmly. In addition, this grasping process should be harmless to the manipulated object, avoiding damage caused by gripper use.
To develop a valid gripper, distinct models must be created and tested. To improve the gripper's capabilities and performance, an iterative approach is used. First, an initial gripper is created. Then, upgrades and refinements are carried out. This procedure is used for the actuation system, mechanical design and other characteristics. Once two models have been created, they are compared using a benchmark testing series. The best model is then selected to ensure improvement. Object grasping tests are performed during the iterative steps to track the gripper's properties during real action tasks.
The main result of this thesis is the creation of a variable-stiffness gripper capable of changing its stiffness at will to adapt to distinct, graspable items. It contains a soft, variable-stiffness core. Testing has revealed its improved performance as compared to its predecessors. The gripper is capable of grasping delicate objects such as fruits, vegetables or eggs without damaging them.
This thesis provides numerous contributions. The most important contribution is the publication of a Utility Model for the final gripper design. This model permits the patenting of this research for future studies and the potential industrialisation of the model.
