**5.1 REMUS UUV and SeaFox USV**

Remote Environmental Monitoring UnitS (REMUS) are UUVs developed by Woods Hole Oceanographic Institute and sold commercially by Hydroid, LLC (Hydroid, 2011). The NPS CAVR owns and operates two REMUS 100 vehicles in support of various navy-sponsored research programs. The REMUS 100 is a modular, 0.2m diameter UUV designed for operations in coastal environments up to 100m deep. Typical configurations measure less than 1.6m in length and weigh less than 45kg, allowing the entire system to be easily transported worldwide and deployed by a two-man team (Fig.10a). Designed primarily for hydrographic surveys, REMUS comes equipped with sidescan sonar and sensors for collecting oceanographic data such as conductivity, temperature, depth or optical backscatter. The REMUS 100 system navigates using a pair of external transponders for long baseline acoustic localization or ultra-short baseline terminal homing, as well as an Acoustic Doppler Current Profiler/Doppler Velocity Log (ADCP/DVL). The ADCP/DVL measures vehicle altitude, ground- or water-relative vehicle velocity, and current velocity profiles in body-fixed coordinates.

Fig. 10. NPS REMUS 100 UUV (a) and FLS arrays (b)

78 Autonomous Underwater Vehicles

; 53 55 52 ( ) *sj j j j*

In this case the last two terms in the joint penalty Δ , similar to that of (27) but developed for

The preceding trajectory generation framework has been implemented on several UMVs. Before presenting simulated and experimental results with specific vehicle platforms at sea, we first introduce two such vehicles in use at CAVR - the REMUS UUV and SeaFox USV. Both vehicles utilize FLS to detect and localize obstacles in their environment and employ the suggested direct method to generate real-time OA trajectories. This section provides system-level descriptions of both platforms including their sensors, and proposes a way of building the OA framework on top of the trajectory generation framework, i.e. enhancing

Remote Environmental Monitoring UnitS (REMUS) are UUVs developed by Woods Hole Oceanographic Institute and sold commercially by Hydroid, LLC (Hydroid, 2011). The NPS CAVR owns and operates two REMUS 100 vehicles in support of various navy-sponsored research programs. The REMUS 100 is a modular, 0.2m diameter UUV designed for operations in coastal environments up to 100m deep. Typical configurations measure less than 1.6m in length and weigh less than 45kg, allowing the entire system to be easily transported worldwide and deployed by a two-man team (Fig.10a). Designed primarily for hydrographic surveys, REMUS comes equipped with sidescan sonar and sensors for collecting oceanographic data such as conductivity, temperature, depth or optical backscatter. The REMUS 100 system navigates using a pair of external transponders for long baseline acoustic localization or ultra-short baseline terminal homing, as well as an Acoustic Doppler Current Profiler/Doppler Velocity Log (ADCP/DVL). The ADCP/DVL measures vehicle altitude, ground- or water-relative vehicle velocity, and current velocity profiles in

 ≤ and *s s* max δ δ≤ .

δ

δ δ

**5. Test vehicles and sensing architecture** 

the architecture of Figs. 2 and 3 even further.

a) b) Fig. 10. NPS REMUS 100 UUV (a) and FLS arrays (b)

**5.1 REMUS UUV and SeaFox USV** 

body-fixed coordinates.

the new controls, enforce *s s*max

1

*q Aw A q B*<sup>−</sup> =− − (32)

To support ongoing CAVR research into sonar-based OA, terrain-relative navigation, and multi-vehicle operations in cluttered environments, each NPS REMUS vehicle has been modified to incorporate a FLS, multi-beam bathymetric sonar, acoustic communications modem, navigation-grade inertial measurement system, and fore/aft horizontal/vertical cross-body thrusters for hovering or precise manoeuvring. (Figure 10b provides a close up of the NPS REMUS FLS arrays with nose cap removed.) To maximize the REMUS system's utility as a research platform, Hydroid developed the RECON communications interface so that sensor and computer payloads can interact with the REMUS autopilot. Using this interface, NPS payloads receive vehicle sensor data and generate autopilot commands based on NPS sonar processing, trajectory generation, and path-following algorithms.

The SeaFox USV was designed and manufactured by Northwind Marine (Seattle, WA) as a remote-controlled platform for intelligence, surveillance, reconnaissance, anti-terrorist force protection, and maritime interdiction operations (Northwind Marine, 2011). SeaFox is a 4.88m long, aluminium, rigid-hulled inflatable boat with a 1.75m beam; 0.25m draft; folddown communications mast; and fully-enclosed electronics and engine compartments. SeaFox's water jet propulsion system is powered by a JP5-fueled, 185-HP V-6 Mercury Racing engine, and can deliver a top speed of 74km/h. Standard sensing systems include three daylight and three low light navigation cameras for remote operation, as well as twin daylight and infrared gyro-stabilized camera turrets for video surveillance. All video is accessible over a wireless network via two onboard video servers.

The NPS SeaFox was modified to enable fully-autonomous operations by integrating a payload computer with the primary autopilot (Fig.11). Meanwhile, the original remote control link was retained to provide an emergency stop function. NPS algorithms running on the payload computer generate rudder and throttle commands that are sent directly to the SeaFox autopilot. Recent navigational upgrades include a satellite compass that uses differential Global Positioning System (GPS) navigation service for accurate heading information, a tactical-grade inertial measurement unit for precise attitude estimation, and an optional ADCP/DVL for water velocity measurements. To support ongoing CAVR research into autonomous riverine navigation, the NPS SeaFox was further upgraded to deploy a retractable, pole-mounted FLS system for underwater obstacle detection and avoidance (Gadre et al., 2009). Figure 12 shows the SeaFox USV operating on a river with its sonar system deployed below the waterline.

#### **5.2 Sonar system**

The NPS REMUS and SeaFox vehicles rely on FLS to detect and localize obstacles in their environment. Both platforms utilize commercial blazed array sonar systems manufactured by BlueView Technologies (BlueView Technologies, 2011). These sonar systems comprise one or more pairs of arrays grouped into sonar "heads." Each sonar head generates a 2D cross-sectional image of the water column in polar coordinates, typically plotted as the image plane's field of view angle vs. range. Due to the sonar arrays' beam width, the resulting FLS imagery has a 12-degree out-of-plane ambiguity. The REMUS FLS system consists of two fixed sonar heads, which provide a 90-degree horizontal field of view (FOV) and a 45-degree vertical FOV. Similarly, the SeaFox FLS system is comprised of twin sonar heads mounted on port and starboard pan/tilt actuators, providing each side with a 45 degree FOV image at an adjustable mounting orientation that can be swept through the water column for increased sensor coverage.

Real-Time Optimal Guidance and Obstacle Avoidance for UMVs 81

Mission goals

**(with Sensors and Controller)**

The proposed OA framework supports both deliberative and reactive obstacle avoidance behaviours. Deliberative OA involves the ability to generate and follow a trajectory that avoids all known obstacles between an arbitrary start location and some desired goal location, whereas reactive OA involves the ability to avoid any previously unknown obstacles detected while following this trajectory. Since the sonar system continuously resamples the environment, this reactive behaviour can be achieved by a deliberative planner as long as i) it executes fast enough to incorporate all new obstacle information from the sonar, and ii) it generates feasible trajectories which begin with the vehicle's current state vector. Specifically, since the REMUS and SeaFox FLS have limited range and limited fields of view in both image planes, new trajectories must be generated continuously (e.g. on some fixed time interval or upon detection of a new obstacle) during execution of the current

As an example of deliberative OA, assume a REMUS vehicle is mapping a minefield with sidescan sonar prior to a mine clearance operation. For this mission, the goal locations are provided by the sequence of waypoints making up a typical lawn-mowing survey pattern. If an obstacle is detected along a specified track line, the preferred OA manoeuvre for this mission would be one that also minimizes the cumulative deviation from this track line, since we desire 100% sensor coverage of the survey area. Hence, deliberative OA implies the optimization of some performance index. Likewise, while digital nautical charts or previous vehicle surveys can be used to identify some obstacles *a priori*, this data is usually incomplete or outdated. Vehicles should be capable of storing in memory the locations of any uncharted obstacles discovered during their mission so that subsequent trajectories can avoid them—even when they are no longer in the sonar's current field of view. Deliberative

Detecting obstacles from sonar imagery is challenging because several factors affect the intensity of sonar reflections off objects in the water column. These factors include the size, material, and geometry of an object relative to the sonar head; interference from other acoustic sensors; and the composition of the acoustic background (e.g. bottom type, amount of sediment, etc.) to name a few (Masek, 2008). Once an obstacle has been detected, other image processing algorithms must measure its size and compute its location within the navigational reference frame. While localizing obstacles via the range and bearing data

**Obstacle Avoidance Framework**

Position Estimate **Augmented Vehicle**

**Environmental Map**

**Localization**

manoeuvre to ensure reactive avoidance of new obstacles.

**5.4 Obstacle detection and mapping** 

OA, therefore, also entails the creation and maintenance of obstacle maps.

Sensor Data

Fig. 13. Components of the NPS OA framework

Reference Trajectory Reference Controls

**Dynamic Trajectory Generator Deliberative**

**Reactive**

Fig. 11. SeaFox sensors and control architecture

Fig. 12. SeaFox USV navigating on the Sacramento River near Rio Vista, CA
