AIRO IDLab Ghent University - imec

In the next decade, robots and other devices that allow us to see, hear, touch, and manipulate objects in places where we are not physically present will play an essential role in our world. While stiff robots have definitely earned their place, especially in manufacturing, current trends indicate that compliant actuators and structural elements are the keys for a safer, more robust, and energy-efficient human-robot interaction.

The current dominant actuation approach in the majority of the robots, both industrial and research, is stiff actuation, which permit precise control of the joints. However, in order to reduce the complexity of the sensing and control modules, and motivated by energy efficiency and a safer robothumanworld interaction, recent trends in robotics tend to use compliant actuation modules that were implemented in state-of-the-art robots. Nowadays, the trend of compliance is extended to other parts of the robot. Flexible materials are now used as structural parts of the robot. At AIRO, we investigate how compliant actuator modules and structures can be integrated into a new robot platform and how they can be used to reduce the complexity of motor control. Our main objective is to, ultimately, derive a set of precise design rules for constructing a well-performing morphology for the task of robot control and cognition. The aim of these new design principles is to develop robots capable of broad classes of motor behaviours without the use of complex control algorithms, which rely on high-quality sensors and accurate models.

From left to right: a robot for weed control, kraken a ball robot and, a the making of a DIY tactile sensor.

Apart from the robot morphology, the sensors also play an important role. At AIRO, we focus on multimodal sensing for perception. Consequently, apart from integrating off-the-shelf camera systems, we also explore ways to enhance robot perception or instrument the learning process. Examples of this are our smarth cloth to instrument the learning of robotic folding, pressure sensor modules in the feet of our quadruped robots, and custom tactile sensors integrated in our lab-made robot grippers.

Over the last decade we have build experience with the design and realisation of cheap compliant quadruped robots. From left to right: Reservoir dog, Oncilla, Spyro 1 & 2, M.A.R.C. and Tigrillo

Publications

Do-it-yourself design for social robots

Vandevelde, Cesar, wyffels, Francis, Vanderborght, Bram, and Saldien, Jelle

IEEE ROBOTICS & AUTOMATION MAGAZINE 2017

Biblio
GECO : the development of a wall-climbing robot

Cardinaels, Rob, Maes, Julie, Deseure, Jordi, Saldien, Jelle, Aelterman, Kevin, Verstockt, Steven, and wyffels, Francis

In Proceedings of CLAWAR 2017 : 20th International Conference on Climbing and Walking Robots and the Support Technologies for Mobile Machines 2017

Biblio
Innovative design of a hexapod scorpion through digital production techniques

Flamand, Stephan, Saldien, Jelle, Deconinck, Pieterjan, wyffels, Francis, Verstockt, Steven, and Terryn, Robbe

In 19th International Conference on Climbing and Walking Robots and Support Technologies for Mobile Machines 2016

Biblio
Kraken: a Modifiable Compliant Robot

Doms, Natan, Dambre, Joni, and wyffels, Francis

2018

Biblio
Quadruped Robots Benefit from Compliant Leg Designs

Willems, Brecht, Degrave, Jonas, Dambre, Joni, and wyffels, Francis

2017

Abs Biblio

A major bottleneck in building autonomous robots is the lack of suitable power supplies. Consequently, researchers aim for more energy-efficient locomotion. One way to achieve this is by using compliant materials in the robot bodies. In this work, we introduce Tigrillo, a small, low-cost quadruped robot platform with modular legs. The robotâs legs consist of a simple two-segmented construction which can be configured stiff or passive compliant. Each leg is actuated at the hip with one motor. By thorough optimization of the open loop control signals we demonstrate that compliance not only results in gaits that are more insensitive to parameter variations, but also leads to more energy-efficient gaits.
Simpler learning of robotic manipulation of clothing by utilizing DIY smart textile technology

Verleysen, Andreas, Holvoet, Thomas, Proesmans, Remko, Den Haese, Cedric, and wyffels, Francis

APPLIED SCIENCES-BASEL 2020

Abs Biblio

Deformable objects such as ropes, wires, and clothing are omnipresent in society and industry but are little researched in robotics research. This is due to the infinite amount of possible state configurations caused by the deformations of the deformable object. Engineered approaches try to cope with this by implementing highly complex operations in order to estimate the state of the deformable object. This complexity can be circumvented by utilizing learning-based approaches, such as reinforcement learning, which can deal with the intrinsic high-dimensional state space of deformable objects. However, the reward function in reinforcement learning needs to measure the state configuration of the highly deformable object. Vision-based reward functions are difficult to implement, given the high dimensionality of the state and complex dynamic behavior. In this work, we propose the consideration of concepts beyond vision and incorporate other modalities which can be extracted from deformable objects. By integrating tactile sensor cells into a textile piece, proprioceptive capabilities are gained that are valuable as they provide a reward function to a reinforcement learning agent. We demonstrate on a low-cost dual robotic arm setup that a physical agent can learn on a single CPU core to fold a rectangular patch of textile in the real world based on a learned reward function from tactile information.

From left to right: a robot for weed control, kraken a ball robot and, a more recent quadruped robot with a flexible spine. The right two photographs are by Guilherme Gerais.