Applications - AUV Development and Testing

The first stage of this process begins with the creation of the vehicle concept, as demonstrated under AUV Configuration. The basic design of the vehicle, the actuators and sensors on board etc., are laid out and assembled on-screen using AUV Sim. Creating sliders linked to the actuators, the vehicle may be test-driven on-screen to check its stability and maneuverability.

Stage 2 - Autopilot configuration
In Stage 2, the design is refined and the autopilots are configured to drive the appropriate actuators. Using the Control Allocation Module built in to AUV Sim, this process is relatively straightforward. The autopilots may then be tuned, either manually or using the self-tuning maneuvers built in to AUV Sim.

At this stage, the vehicle may be tested in simulation, running through the mission profiles and tasks which are to be performed by the real vehicle, to ensure that the vehicle design meets the required criteria in terms of stability, speed, maneuverability and controllability.

Stage 3 consists in downloading the vehicle configuration parameters to the 3D-MC Autopilot board for the first stage of hardware testing. The 3D-MC board is connected to a vehicle simulator (AUV Sim) via serial ports (for most navigational instruments) and analog and digital channels as appropriate. For example, the depth sensor may generate a voltage proportional to pressure: for testing, this voltage is generated by the AUV Sim via a PC Analog Output card.

To perform the test, using this configuration, the 3D-MC Autopilot board receives commands from the Remote Control Workstation - which is a PC, running AUV Sim in External Vehicle mode. At the same time, the 3D-MC board may send sensor and machinery data back to the Remote Controller, enabling the operator to monitor the vehicle and to send mission commands down to it. This stage provides a good measure of operator training.

In Stage 4 the vehicle actuators may be incorporated into the test. This allows the system to be tested, once more, under simulated conditions, but the test may include more of the actual vehicle hardware.

One crucial aspect of this test is that the actual machinery lags and response characteristics may be included in the test - hitherto tested using the lag models within AUV Sim. This step ensures that the machinery performance is adequate for system sea trials. Most importantly, this test may be used to ensure that actuators are accurately controlled. For example, it is important for optimum performance, that hydroplanes reach the setpoints commanded by the controller, to a good degree of accuracy - backlash in a spindle mechanism, for example, may seriously degrade the controller performance. If carefully performed, this pre-trial test may be used to ensure that all such problems are eliminated.

In this way, before committing the vehicle to the water, the entire system, including the on-board autopilot, can be tested by performing a virtual sea trial. This is a time and cost effective way to eradicate any residual problems and to prove the performance of the complete system. Full mission testing can be carried out at this stage to check that the mission is viable. This is shown schematically above.

Finally, the vehicle is committed to the water for Sea Trials. Again, the controller on board is connected to the Remote Control Workstation via an umbilical cable or an acoustic modem link - or, in the case of a surface vehicle, a radio modem, for monitoring, mission download, and intervention as necessary. Gradually the degree of intervention may be reduced until the vehicle is operating in fully autonomous mode.

On completion of Sea Trials, the vehicle is commissioned ready for missions.