Professor Althoefer is an experienced roboticist leading competitively-funded research on soft robotics, intelligent micro-sensing systems and interaction dynamics modelling with applications in minimally invasive surgery, assistive technologies and human-robot interaction at Queen Mary University of London, acquired more than £5.5M as PI from national/international funding bodies and successfully completed 21 PhD projects.
Professor Althoefer's research team currently comprising 10 postdoctoral research associates and PhD students is involved in funded collaborative research with leading London hospitals and European research organisations creating novel robot-assisted solutions for ergonomically-optimised human-robot interaction, intelligent manipulation based on embedded force and tactile sensing and novel stiffness-controllable robot structures.
Over the last decade, the team has built a large portfolio of projects in application-oriented research for the healthcare and manufacturing sectors with funding from organisations such as EPSRC, European Commission (including coordination of two EU-projects), Wellcome Trust and UK-based charities, exceeding £30M and producing more than 250 peer-reviewed papers.
The field of human-robot interaction (HRI) has seen a dramatic development over the last decade, with a wide range of researchers moving the field forward considerable. HRI has gained in importance with an observed need for robots that can operate in the vicinity of humans or even get in contact with them. Soft material robotics has a lot to offer in this context. Because of their compliance, robots made from soft materials are considered much safer than their rigid component counterparts.
There is a general consensus that pneumatically or hydraulically actuated robot arms made from soft materials have low computational requirements concerning the computation of the interaction forces when in contact with their environment. Their inherent compliance makes them suitable especially for tasks that require the handling and manipulation of fragile objects. The robot arms adapt and conform to the objects that they are in contact with and the interaction forces can be determined as an easy-to-compute function of the pressures in the robot's actuation chambers.
This simplification in the way objects are handled owing to the robot's structure and material properties is considered an example of morphological computation. Due to the highly nonlinear behaviour of the materials used, such as silicone rubber or tailored textiles stretched by compressed air, the relationship between the robot's pose and the control commands (actuator pressures) are difficult to describe using analytical models. Employing deep learning methods, the relationship between the control commands and the pose of the robot structure in the context of a possibly highly dynamic scenario can be determined.