Unmanned Systems Technology 007 | UMEX 2016 report | Navya ARMA | Launch & recovery systems | AIE 225CS | AUVs | Electric motors | Lethal autonomous weapons

67 AUV that can follow underwater pipelines and even turn their valves on and off. The Girona 500 AUV has been fitted with an ‘end effector’ to turn the valve, along with Reactive Fuzzy Decision Maker (RFDM) software to evaluate the safety of turning the valve. The end effector has been designed in three parts. A passive gripper grasps the valve handle while a camera inside the end effector monitors the valve wheel, and a torque sensor evaluates the quality of the grasping and the torque needed to turn the valve. The RFDM combines the data from the torque sensor and camera so that the end effector does not apply too much force. This required a change to the original single thruster design of the AUV, as an additional external thruster had to be installed to keep the AUV in the right position while it turned the wheel of the valve. New approaches A new technique for monitoring is to use swarms of smaller AUVs to cover a wide area. One such system is the Comet platform, developed by RTsys of France, which weighs less than 35 kg and can travel at up to 15 knots. It can also cover 200 km at 3 knots on a single deployment from its rechargeable batteries. It uses between three and ten collaborative AUVs operating simultaneously at distances of up to 1500 m apart, communicating using RTsys autonomous modules that can send and receive voice and data across more than 2000 m. The modules can also work as a digital audio player and recorder controlled by underwater acoustic comms. The platform, launched in March 2016, is aimed at shallow water and harbour environments, and is designed for harbour security, discreet underwater listening, and underwater control and telemetry, and has already been demonstrated operating in swarm mode scouting in front of a vessel. AUV design is also opening up new areas of operation. The Caddy research programme funded by the EU, for example, aims to develop an AUV that can act as a ‘buddy’ for a diver. The system consists of an AUV linked to an autonomous surface vehicle to monitor and protect a diver operating on their own. There are three elements to the project. The ‘Seeing the Diver’ research theme focuses on a 3D reconstruction of the diver, including an estimation of their pose as well as recognising hand gestures. The ‘Understanding the Diver’ element focuses on adaptive interpretation of the data about the diver’s pose and physiological measurements in order to determine the state of the diver, while the ‘Diver-Robot Cooperation and Control’ theme uses the recognition of the hand gestures of the diver to control the AUV. To achieve this, new modems and navigation systems were developed by one of the research partners, the University of Newcastle in the UK. The system operates at only 100 bit/s but provides reliable acoustic comms between diver, buddy and surface vehicle. The first AUV (called, unsurprisingly, Buddy) has been built by the project’s lead partner the University of Zagreb, and has six thrusters, stereo and mono cameras, high-resolution multi-beam sonar and a mounted underwater tablet PC for providing feedback to the diver. The Zagreb team has also modified diver symbolic language to create Caddian, the control language for the system. More than 40 commands and messages have been defined that can be recognised by the AUV for control and comms. Conclusion As AUVs operate in deeper waters from 4000 to 6000 m, so the need for autonomous operation rather than remote operation is driving research into new control and sensor fusion techniques. Borrowing research from autonomous space systems is allowing AUVs to set their own courses and travel autonomously, while combining sensor data with autonomous control algorithms is allowing AUVs to track shoals of marine life to monitor changes in the sea’s ecosystem over time. Unmanned Systems Technology | April/May 2016 The Caddy research programme aims to develop an AUV that can act as a ‘buddy’ for a diver Using a reference model of the pipeline, a new survey would be able to detect any movement of the pipeline, such as over-bending