Aerial Swarms Workshop at IEEE/RSJ IROS 2019

© 2019 EPFL

© 2019 EPFL

LIS organized a workshop about Aerial Swarms at the IROS 2019 conference. The workshop was really successful with fruitful discussions and an average attendance of 80 people with peaks of up to 120 people. The aim of this workshop was to trigger a scientific discussion on the most recent advances and challenges of the science and technology of aerial swarms.


Recent advances in sensor technologies have greatly enhanced the capabilities of unmanned aerial vehicles (UAVs). One consequence of this has been the growing interest in multi-aerial robotic systems often simply called "aerial swarms". The interest in single aerial robotic systems is still bustling and attracting many researchers worldwide. However, in the last decades, aerial swarm robotics consolidated itself as one of the most challenging, exciting and multidisciplinary fields of robotics. These systems have many advantages with respect to single-robot systems. In particular, aerial swarms have the possibility of fulfilling more complex tasks faster, and they are expected to be more robust to failures. Another advantage is that it is possible to share capabilities over the group (e.g., one part of the group is the eyes of the group, one part is the hands of the group). Real-world applications that could benefit from the use of aerial swarms include patrolling, exploration, search and rescue in large areas, as well as cooperative transportation and construction.

Workshop Goals

This workshop brought together the subsets of the two communities of multi-robot and bio-inspired swarms dealing specifically with systems of multiple aerial robots. The workshop aimed at triggerring a scientific discussion on the most recent advances and challenges of the science and technology of aerial swarms. We believe that the two aforementioned communities tackle remarkably similar problems but in different manners. The common problems they try to focus on and solve are the ones of perception, estimation, control, and navigation of groups of UAVs in different environments. The multi-robot community often wants to ensure a particular configuration and performance tackling, for example, the swarming problem with tools from algebraic graph theory. Some of the works assume that the robots know their absolute positions and many assume that a robot can communicate with other members of the group. The bio-inspired community tries to tackle the same problem taking inspiration purely from biological systems. Therefore, they usually assume neither communication nor knowledge of the absolute positions of the agents. However, this community seldom implements the devised algorithms on real robotics systems to verify their applicability in the real world.

In general, we believe that there is not a gap between the two communities, but with this workshop we tried to highlight the differences in their approaches. We reached our goal of sharing knowledge and to understand the main directions towards which the two communities are heading.