DCSC Robotics lab

Robotics platforms

We have developed two power and computational autonomous legged platforms inspired by RHex: the Zebro, illustrated in Figure 1 left, and RQuad, illustrated in Figure 1 right. The Zebro Light robot (see Figure 2) was developed by a group of BsC students.

Figure 1: High performance power and computation autonomous robots

Figure 2: Zebro light – Low cost, low weight, high power robot

Relevant publications:

  • G.A.D. Lopes, B. Kersbergen, T.J.J. van den Boom, B. De Schutter, and R. Babuska. Modeling and Control of Legged Locomotion via Switching Max-Plus Models. Accepted to IEEE Transactions on Robotics.

  • G. A. D. Lopes, B. Kersbergen, B. De Schutter, T. J. J. van den Boom, R. Babuska. Synchronization of a class of cyclic discrete-event systems describing legged locomotion. Submitted. Available at arXiv.

  • G.A.D. Lopes, B. De Schutter, and T. van den Boom. On the synchronization of cyclic discrete-event systems. IEEE Conference on Decision and Control, 2012

  • F. Zhang, G.A.D. Lopes, and R. Babuska. Stiffness and Damping Scheduling for Legged Locomotion. Accepted to IEEE International Conference on Robotics and Biomimetics, 2013.

  • D. van der Lijn, G.A.D. Lopes and R. Babuska. Motion Estimation Based on Predator/Prey Vision. IEEE/RSJ International Conference on Intelligent Robots and Systems, 2010

Mechanical design

Towards improving the performance of the legged robotic platforms, we have developed hardware elements such as the novel leg-wheel-fin deformable legs prototypes illustrated in Figure 3. The leg illustrated in Figure 3 left is equipped with a FBG graded fiber optical sensor for direct deformation measurements. Figure 4 left illustrates a clutch-based leg-wheel concept driven by a single actuator developed by a group of BSc students. Figure 4 right illustrates a light weight 4-link manipulator developed for education and research (extensible to n-links).

Figure 3: Novel deformable leg-wheel-fin designs for search and rescue robotic applications

Figure 4: Left:Leg-wheel concept with automatic clutch-based mode switching. Right: Modular n-link robot arm for education and research. Lightweight with force sensing.

Relevant publications

  • G. A. D. Lopes and F. Zhang. Design and sensing of a flexible robot leg. Presented at IEEE/RSJ International Conference on Intelligent Robots and Systems, 2013

Electronics design

Our team has developed a new class of high-power low-size low-latency modular motor control boards for applications in robotics. Figure 5 illustrates 3 boards developed in our lab, together with a Raspberry Pi board illustrated for size comparison.

Figure 5: High performance motor control boards. Top left: ‘‘hip board’’ drives a motor with 40A max current and 5A nominal current, 24V. This board is embedded with 2 encoder counters, temperature and current sensors. Bottom left: the ‘‘spine board’’ communicates with up to 16 hip boards in parallel at very low latency. The spine board communicates with off-the-shelf CPU's (see top right image, a Raspberry Pi board) using the USB bus, enabling control rates up to 1 KHz. Bottom right: is a power board for battery driven autonomous systems that enables hot-swapping of batteries, total voltage and current measurements, and circuit protection.

Software

Figure 6 illustrates a novel type of web-based real-time graphical user interface (GUI) that leverages on an existing multitude of freely available libraries (e.g. google maps) to create rich interfaces that can run on desktop computers or mobile devices without the need to install specific/native software or any browser plug-ins.

Web-based real-time graphical user interface for robotics. Includes 3D rendering, maps, FSM, controller buttons, console, graph plotting, etc.