During situations such as disaster response, such as flooding or landslides, there is at times need for temporary structures such as bridges and landing platforms so that heavy machinery may reach an area to provide assistance.
How to create that platform in a tricky situation in a way that minimises further risk to human life?
A modular drone that can form a larger structure with other drones in mid air has been created by scientists at the school of Engineering and Applied Sciences at the University of Pennsylvania that may address this need.
In a video published on YouTube by Vijay Kumar Lab, two drones are shown encased in a box shaped frame as they fly towards each other in a three-dimensional environment. Each module consists of a quadrotor platform and a docking method that allows the pair of structures to self-assemble in mid air.
Created in partnership with researchers from the Modular Robotics Lab of University of Pennsylvania, a paper titled “A Decentralized Algorithm for Assembling Structures with Modular Robots” discusses a system that mimics the construction techniques of ants. When ants build a nest they build temporary bridges to join areas so that they may transport supplies to otherwise inaccessible areas.
Kumar and his team have proposed that by mimicking these techniques using an ant-inspired algorithm they could also create autonomous robots that can self-assemble and thereby create temporary structures that could be used, for example, in times of flood or other extreme events.
The researchers have dubbed these drones ‘ModQuads’ and in the video they demonstrate first two, then four, then six of the devices joining together to create a hovering platform.
Sounds like floating platforms right out of Mario World to us!
University of Pennsylvania is a hotbed of innovative aerial drone development it seems. Earlier this year they also demonstrated a swarm of robust aerial robots that could fly much like a large flock of birds or bees. We’re all familiar with collision avoidance technology that is present in most present-day drones.
The technology that applies collision avoidance abilities in unmanned aerial vehicles involves sensors measuring proximity to their neighbours and adjusting flight paths accordingly. However when flying in large numbers the complexity of these calculations scale exponentially.
Penn Engineering in collaboration with the GRASP Laboratory demonstrated in another video published last month another drone design that that incorporates the capability of sustaining a certain amount of force and controllers that can recover from trajectories that have been diverted by collisions. By doing this, large swarms of drones are then able to follow a predetermined course without complete and accurate data about their environment.