Robot - Sensor Network Interaction

Our application focus to date has been on monitoring marine environments and inland waterways where we have leveraged our expertise in underwater robotics - particularly our award winning Starbug autonomous miniature submarine. We are also exploring the interaction between sensor networks and autonomous ground vehicles, as well as unmanned aircraft systems.

Application Areas

Monitoring Waterways and Marine Environments

Many scientifically important marine and inland water environments are too large and dynamic for efficient and cost effective data collection by robotic systems alone. A method for improving capture of relevant dynamic properties in these environments is to integrate marine sensor networks and Autonomous Underwater Vehicles (AUVs) and Autonomous Surface Vehicles (ASVs) in a way that they complement each other.

Using integrated sensor networks comprising of static sensor nodes, Autonomous Underwater Vehicles (AUVs) and Autonomous Surface Vehicles (ASVs), allows scientists to develop an overall picture of the status of waterways and marine environments.

We have focused on developing solutions that allow robots and sensor networks to effectively interact in an effort to monitor particular hydrodynamic and biogeochemical events in real-time without human intervention. Our research in this area covers in-network event detection, distributed communication and data sharing, network calibration, multirobot interaction, model-based path planning and resource management.

Read more about our use of sensor networks for water quality monitoring.

[Top]

Research Topics

Floating sensor node and autonomous catamaran on Lake Wivenhoe, South East Queensland.

Integration Frameworks

We have developed a suite of software architectures and libraries to allow rapid development of integrated robotic and sensor network systems. Frameworks include the Dynamic Data eXchange (DDX) used on all our robotic platforms, and the Fleck Operating System (FOS) used on our embedded sensor network hardware. Current research is investigating different programming paradigms to allow operators and robots to request services from the collection of deployed resources in a manner that is transparent to the sensor network.

Data Muling

In remote environments, sensor nodes can become isolated when their communication distance is too great, when hardware fails, or due to environmental factors. In these situations, the sensor nodes can store measurement data onboard until a mobile agent can retrieve this information for return back to base. We have developed and demonstrated various techniques for reliable identification of sensor nodes by robotic systems, interaction between the robot and sensor node, and data retrieval by the robot. We have performed data retrieval experiments in disconnected land-based and underwater sensor networks.

Starbug autonomous submarines. By sharing information, robots can expand their knowledge horizons.

Multirobot cooperation

Sensor networks can provide a communication framework to allow multiple robotic systems to interact and achieve complex tasks. They can also offer information to, or request the services of the robots to improve the scientific understanding of various processes within the environment, such as detailed sampling of events, or improve the operational functionality of the deployed resources. Our research in this area covers multirobot interaction with time-delayed communications, cooperative and adaptive mapping, and automated robot deployment and retrieval of nodes.

[Top]

Key Staff

Dr Matt Dunbabin

Dr Ash Tews

Dr Wen Hu

Dr Robert Zlot

[Top]

Publications