SeaSlug: A Low-cost, Long-duration Mobile Marine Sensor Platform for Flexible Data-collection Deployments

B. Mairs, R. Curry, G. Elkaim

Abstract:	Earth is a planet whose weather and ecosystems are heavily influenced by the oceans that cover 71% of its surface. 95% of the ocean is estimated as unexplored and so human understanding of its influence is severely lacking. The sea-surface microlayer, corresponding to the top 1m of ocean, is known to be an important factor in its influence as it affects algal blooms, ocean acidification, and global climate change among many other as-of-yet unknown biological and chemical processes. And while many tools are available for collecting data about the ocean's surface, currently only manned surface vessels offer the necessary flexibility for general-purpose oceanographic data collection. These vessels have significant limitations, however, in that they pollute heavily, are often difficult to obtain, and costs are high. Autonomous surface vessels have emerged as possible replacement for manned vessels as a remedy for these problems, but none have the necessary combination of sensor flexibility and long-term endurance. Additionally the cost is often exorbitant those platforms that are available, which limits their use to only well-endowed organizations. If this cost can be reduced, the amount of data collected, currently a large limiting factor in oceanography, could be significantly increased. For this reason the SeaSlug, a new autonomous surface vessel for marine sensing applications, was created. Its design goals include a low total and operational cost, being sensor-agnostic, and endurance for continual deployments of up to several days. It was built using a custom mono-hull design with electrical propulsion. With self-righting capabilities, this design is more robust than either catamaran or wind-powered vessels and allows for a large solar-cell array on the deck. It's low beam and low-speed propeller facilitate high endurance while still allowing for a peak speed of 2.3m/s. Available power is broken into a 12V power rail for electronics and a 24V rail for the actuators. 1kW of solar charging feeds into the 7.6kWh of onboard lead-acid batteries. Even with these substantial capabilities, the estimated cost of the SeaSlug is under $30,000. An onboard autonomous controller provides waypoint following utilizing an adaption of the L2+ control law from aerial vehicles. By including a model of the vessel kinematics, the commanded lateral acceleration from the L2+ controller can be mapped into a commanded rudder angle suitable for the SeaSlug. This control is an improvement upon earlier path-following algorithms that rely on a look-ahead vector. It also includes several modifications designed to improve stability in operational environments that may experience significant external forces. The onboard controller relies on the simple inverse-bicycle model, common to all rear-steering surface vehicles. Its use of only two parameters facilitates estimation of these parameters from test data. While simple, the use of this model does not severely impact the performance of the controller. Both water speed and wind speed sensors are available and integrated as part of this model further improving position estimation. The operator interface to the vessel for mission retargeting, altering parameters, and telemetry capture is provided by the open-source QGroundControl (QGC) project and it's MAVLink protocol. Manual control is transmit to the vessel through the groundstation interface via a USB joystick (a backup controller using a direct connection to the actuators is provided with a standard RC transmitter). All onboard telemetry is captured to a datalogger aboard the vessel, but is also streamed at a lower datarate to the groundstation for immediate analysis. The vessel's waypoint list can be viewed and altered graphically using a Google Maps-like interface. Parameters affecting almost all of the vessels capabilities can also be altered and tuned during deployments. Testing of the SeaSlug follows a progression through four separate stages: full simulation, hardware-in-the-loop (HIL), HIL with actuator feedback from small-scale subsystems, and onboard dry-dock testing. The first simulation step is performed entirely on a PC and runs the same controller algorithms as those onboard the vessel. The next stage of testing involves hardware-in-the-loop (HIL) using the hardware controller board that controls the SeaSlug. The simulator then runs only the environmental portion of the simulation with the controller operating as if it was aboard the vessel. The final HIL testing mode includes using the actual actuator sensors, with feedback from them feeding back into the simulator. This allows for testing of the main controller with real actuator dynamics. In this mode the rudder is operated identically to how it would be during a live test. Tests in the Santa Cruz harbor and in the Monterey Bay demonstrate the capabilities of the performance of the onboard control algorithm and accuracy of the vehicle model. These tests show that the L2+ control algorithm is capable of commanding an RMS cross-track error of 1m and heading error of 2.6 degrees along straight-line segments. Additionally its endurance was under analysis during theses tests. Power draw without sensors at a constant speed of 2m/s is 600W over the course of an entire 4-hour test deployment. With onboard power storage at 7.6kWh, 12 hour deployments are within the safety margin of the vessel without factoring in solar charging capabilities. With those deployments are on the order of days with the vessel easily able to run through the night.
Published in:	Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013) September 16 - 20, 2013 Nashville Convention Center, Nashville, Tennessee Nashville, TN
Pages:	605 - 611
Cite this article:	Mairs, B., Curry, R., Elkaim, G., "SeaSlug: A Low-cost, Long-duration Mobile Marine Sensor Platform for Flexible Data-collection Deployments," Proceedings of the 26th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS+ 2013), Nashville, TN, September 2013, pp. 605-611.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In