STARFISH

Small Team of Autonomous Robotic "Fish"

AUVs provide an effective way to survey and monitor our oceans. As opposed to large, complex and expensive AUV that carries all types of payloads in one single body, our research is focused on developing a heterogeneous team of small, low-cost AUVs that can cooperate and perform complex tasks. The research in this area spans many enabling technologies across disciplines and various types of autonomous vehicles. AUV research is one of the key focus areas for ARL, and in this effort we collaborate with many groups in NUS, NTU, MIT, etc to try and achieve this vision of teams of cooperative low-cost vehicles.

The STARFISH AUV

As part of the vision of STARFISH, a detailed specification of a highly modular, open-architecture STARFISH AUV capable of being a research platform to test collaborative AUV algorithms was finalized in October 2006. The design allows us to easily add and move modules in the vehicle and alter the sensing and actuation capabilities.

Bluestar Prototype

The first STARFISH AUV prototype, codenamed Bluestar, was fully integrated at the end of 2007. The Bluestar prototype weighs less than 45 kg and is about 1.6 m in length (excluding payload), and consist of a number of base sections. The base sections provide basic positioning, navigation, obstacle avoidance and communication capability that is common to all STARFISH AUVs. An advanced navigational payload section with a DVL has been developed for this prototype to provide accurate underwater positioning. The modularity of the system allows us to continuously improve the system design with minimal down-time from field trials.

Bluestar prototype at sea

Mark II Prototype

A second prototype (yet to be christened) is scheduled to be completely tested at sea and perform cooperative missions along with Bluestar by end of 2009. This prototype is built with improved robustness, higher degree of modularity, and ease of manufacturing. The aim is to replicate this design to form a larger STARFISH AUV team in the next year.

Bluestar (back) and the Mark II prototype (front)

Software and Hardware Modularity

STARFISH is designed with software and hardware modularity in mind. The hardware modularity is provided through clear definition of simple standardized electrical interfaces and mechanical interfaces. The software modularity is achieved through the DSAAV architecture. At a command & control level, the AUV is controlled by a set of interacting intelligent software agents or immobots.

Payload Sections

Additional sections can be built based on payload specifications that will be made publicly available shortly; these modules increase the sensing or actuation capability of the STARFISH AUV. There are two in-house payloads currently built, a DVL advanced navigation payload and a sidescan sonar payload, while another towed array receiver payload is currently under development. A group of MIT researchers are currently developing another STARFISH sensor payload to detect hydrocarbons, bio-entities, etc using laser-induced fluorescence and LED-induced fluorescence. A mass spectrometry payload to detect low molecular weight hydrocarbons is also in the pipeline. In addition to sensor payload sections, we are currently considering actuation payload sections to add bow thrusters and buoyancy pumps with an aim to provide hovering capability to the AUV.

Sidescan Payload Section

Field Trials

The Bluestar prototype has gone through extensive confined water trials and a number of open sea trials in local Singapore waters. The trials normally consist of a number of about runs, each up to several km in length, with a mix of surface and dive mission points or objectives. Typical positioning error at the end of the run is less than 1% of the total distance traveled, when the AUV is used with the advanced navigation payload.

The enhanced-Folaga (eFolaga)

The Folaga is a low-cost hybrid glider-like AUV developed through a partnership between ISME, NURC and GraalTech. Through a collaborative project, we are working with them to develop an enhanced eFolaga that is modular and can operate as part of our STARFISH team of AUVs. We envision many disparate low-cost AUVs to cooperate towards a mission goal; we will be able to test some of these ideas with the STARFISH AUV and the eFolaga.

Launch and recovery of AUVs

Autonomous launch and recovery of AUVs is still an unsolved problem. In collaboration with NURC, we recently proposed an autonomous initial capture system for AUV recovery [1]. This system aims to form an important component of any autonomous launch and recovery system for AUVs. We are now in the process of developing a prototype to test the idea.

Applications of AUV teams

Networked Environmental Monitoring using AUVs

In collaboration with MIT, under the CENSAM program, we are exploring cooperative environmental monitoring using our STARFISH AUVs and other sensing platforms. Fixed platforms such as PANDAs and surface platforms such as autonomous kayaks will be used in this project as communication and navigation gateways or as sensor nodes. Specially designed sensor and actuator payloads on STARFISH AUVs will allow environmental surveys to be performed quickly and efficiently. The project started in early 2008 and will last until 2012.

Off-shore surveying using AUVs

Typical off-shore and seismic surveys today are very expensive in terms of money, logistics as well as environmental impact. As oil and gas industry moves towards exploitation of ultra-deep reservoir, higher acoustic energy is required when surveying from a surface vessel, an approach unfriendly to cost and environmental. The role of AUVs in oil and gas industry has started to gain importance since the mid-1990s and AUVs like the HUGIN 3000 are now well established as accurate and efficient tools for certain kind of surveys. We are assessing survey techniques using a fleet of collaborative lower cost AUVs to survey an area of interest. If successful, the potential benefits that these techniques may bring include lower cost, higher resolution for surveys, better coverage and lower environmental impact.

Relevant Publications