This workshop focuses on technologies that allow robots to be designed by software. We aim to generate discussion into the design of bespoke specialized robotic systems that are poised to provide superior performance compared to currently-available options. In such an approach, robots may optimize their hardware (morphologies, sensor/actuator configurations, and the materials they are made from), and software (controllers). In each case the onus is on techniques that automatically design and create robots, or robot components, with no (or minimal) human intervention. Such an approach is becoming increasingly feasible, due to the rapidly-advancing push in manufacturing technologies, including 3D printing & high-throughput materials characterization.
We seek participation from diverse fields, including robotics, mechatronics, materials characterization, structural design, and controller optimization, to overview the current state of the art, and to propose new directions.
Confirmed Invited Speakers
. Automatic generation of bodies and controllers
. Evolutionary and machine learning approaches
. Designing materials for robots, characterization, high-throughput techniques
. Limits of morphological complexity, designing for morphological computing
. Use of Additive Manufacturing, printable components, new sensors, actuators, and robot materials. Online/offline environmental adaptation, shape-change
. Autonomous evaluation/validation techniques
. Combining simulation and physical tests, surrogate modelling
. Removing humans from the design and construction process
Submissions may be accepted as oral presentations or posters. All submissions must be original and not previously or simultaneously submitted to another venue.