- Creation: 01/30/2020
- Update: 12/01/2020
Our project aims to create a human hand prothesis which can reproduce all the moves a human hand can do in an autonomous way. We are a team of 8 students from the DaVinciBot association, the association that finances the entire project.
Since servomotors, or hard robotics in general, is the most spoke about technology in this field, we chose to focus on soft robotics, which seems to be a promising innovation in the scientific world. Our goal is for our members to improve their knowledge about soft technologies such as pneumatic or hydraulic (moulding of the silicone materials: tutorial to come), or shape memory materials. We also aim to improve our programming and electronic skills that will help us with the system subservience and MachineLearning.
In a more global way, this project represents the first extensive project from the Health Department, which is an area for the future, playing with the notion of transhumanism, a subject that will inevitably be largely studied in the coming years.
We spent the first year discovering soft robotics and making first attempts at using them on prototypes. We made a few of them and ended up with a semi-rigid prototype (rigid for the palm, soft for the fingers’ flexion and spacing). After these research and testing phases, we decided to use exclusively pneumatic technology.
We now are in the second year of the project and have enough knowledge about pneumatic systems. We were also able to improve our skills on the moulding thanks to a workshop lead by Yoav Reches, co-founder of Formlabs. We decided to go for an ambitious take: an entirely flexible hand, made of elastomer, with an integrated pneumatic muscular system. This idea is still in the modeling process and needs the use of sacrificial moulds, a method we are experimenting (tutorial coming soon).
As of the programming part, we are getting used to the use of diaphragm pumps and electro valves (to direct the different air flows) by making a control board.
• Mould release of a first elastomer membrane with integrated muscles, in a shape of a slip-on mitten.
• Full control over the board to allow us to pilot the mitten.
• Start studying MachineLearning (or DeepLearning). It will allow the hand to autonomously grab an object according to the print left on the palm by the object. We are working with Brice Parilusyan on this. He is a fifth year ESILV student, working on an artificial skin that can be used as a surface pressure sensor.