Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
The rapid evolution in microsystems and optoelectronics creates the demand for robotic production means with increased performance. The required precision for specific tasks in machining, positioning and assembly is a factor 10 to 100 better than 1 micrometer, while keeping high dynamic properties and motion ranges of up to 1 centimeter. Every application has its own constraints, for the layout of the desired degrees of freedom of the robot as well as for the available volume, and induces a considerable development duration. This thesis aims at making the first phase of manipulator design process more efficient, from the specifications gathering down to the kinematics structure design. The solutions space of all possible kinematics in 3D being incredibly broad and rich, several simplifications have been introduced, in order to respond to the designer's need for structure and pragmatism. On the one hand, only structures having forces aligned with the main orthogonal axes in space are considered, which allows a systematic classification of the kinematics chains and allows us to define the generic concept of "functional" joint with well defined kinematics properties. On the other hand, the proposed methodology is limited to qualitative aspects of motion, for small motion ranges, with a systematic use of the "parasitic" motion concept. The first phase of the design problem is solved in two steps: first, the "elementary" kinematics function of the robot is identified, and second, alternative structures are generated in a creative process, following a set of principles and using a variety of partial and global solutions. The generation of structures with parallel or hybrid kinematics is thereby greatly enhanced and possible problems of mechanical overconstraint are systematically identified. Elastic solid state joints are incorporated in the generated structures on the one hand through a set of design recommendations and on the other hand through specific substitutions. A 6 degrees of freedom micromanipulator with high dynamics and nanometric position resolution was designed during this thesis and a patent application was filed. The prototype "Sigma 6" was built and successfully used to perform an active fiber alignment procedure. A new family of rotational structures with 1 or 3 degrees of freedom, a remote rotation center and reduced parasitic rotation center shift was designed. A new decoupled arm with increased mechanical load capability made of 18 elastic joints was designed and a prototype built, as an alternative to traditional arms ended by a universal and a spherical joint.
Aude Billard, Mikhail Koptev, Nadia Barbara Figueroa Fernandez
,
Auke Ijspeert, Karen Ann J Mulleners, Kamilo Andres Melo Becerra, Alexandros Anastasiadis, Laura Isabel Paez Coy, Eric Daniel Tytell