Distributed system of autonomous buoys for scalable deployment and monitoring of large waterbodies

Abstract

The design, construction, and testing of a large distributed system of novel, small, low-cost, autonomous surface vehicles in the form of self-propelled buoys capable of operating in open waters is reported. We detail the successful testing of collective behaviors of systems with up to 50 buoys, achieving scalable deployment and dynamic monitoring in unstructured environments. This constitutes the largest distributed multi-robot system of its kind reported to date. We confirm the robustness of the system to the loss of multiple units for different collective behaviors such as flocking, navigation, and area coverage. For dynamic area monitoring, we introduce a new metric to quantify coverage effectiveness. Our system exhibits near optimal scalability for fixed target areas and a high degree of flexibility when the shape of the target changes with time. This system demonstrates the potential of distributed multi-robot systems for the pervasive and persistent monitoring of coastal and inland water environments.

Gradual Collective Upgrade of a Swarm of Autonomous Buoys for Dynamic Ocean Monitoring

“…In [12], we presented a study of dynamic ocean monitoring using a homogeneous swarm of 50 buoys. Such decentralized and cooperative systems primarily owe their outstanding effectiveness to the large number of agents put together: a greater number of agents allows for vaster waterbody coverage and/or refined multi-point sensing.…”

“…On the other hand, the new hull has a larger packing density, which facilitates stackability and transportation of large numbers of platforms to remote field areas (see Fig. 1 in [12]). The mechanical design of v2 and the distribution of components inside the main hull has been designed and tested to not compromise the hydrostatics of v1, thereby retaining the critical self-righting feature of the v1 design.…”

“…Our multi-robot system can perform a range of collective behaviors achieved by means of a cooperative control strategy supported by distributed communications as described in full details in [12]. For instance, this distributed robotic system can perform a number of elementary swarming behaviors, including flocking, navigation, and area coverage.…”

“…To quantify the response of the system, we consider two metrics for the dynamic coverage performance that were previously used when testing with a homogeneous swarm [12]. The first one is the "tessellation performance" P T , defined as the inverse of the relative size of the largest cell assigned to any agent after segmenting the target area with a Voronoi tessellation, or…”

Decentralized Multi-Floor Exploration by a Swarm of Miniature Robots Teaming with Wall-Climbing Units

“…For example in [9], a decentralized algorithm was used to localize a flock of robotic sensor networks. Environmental monitoring tasks were performed by a decentralized swarm of robots in [10]- [12], including with heterogeneous swarms [13]. Decentralized exploration and mapping of unknown indoor entity was demonstrated with a pair of autonomous quadcopters in [14].…”