Team Members

For the past year the University of Rhode Island's Autonomous Underwater Vehicle (AUV) Team has been working to develop an AUV capable of satisfying the mission requirements as stated by AUVSI and ONR's AUV 7 Annual Competition Rules. The vehicle, named Prowler V, is a continuation of the vehicle developed and modified over the past 4 years. The Prowler V offers increased processing and power efficiency, durability and reliability. Major modifications to this year's vehicle include a more adaptable and durable chassis, the replacement of the control plane and thruster motors with new efficient servos as well as a completely refurbished AUV control system. This report outlines the details of how this AUV operates and the systems it is composed of. Introduction As the University of Rhode Island enters its fifth year in the International AUVSI and ONR's Autonomous Underwater Vehicle Competition, the Prowler is entering a new age in development. The past several years of experience have led to a sophisticated yet lean autonomous underwater vehicle (AUV). The strategy for developing the Prowler for this year’s competition followed the same guidelines used in the previous vehicle designs: efficiency and simplicity. The Prowler V, shown in Figure 1, is capable of making fully autonomous, yet predictable, decisions using a simple and efficient network of microcontrollers. The AUV also has more awareness of its surroundings than ever before using a variety of acoustic, optical, pressure and magnetic sensor systems. Figure 1 Design drawings of the Prowler V. The mission can be separated into 3 stages. The first stage is to pass through the validation gate, something that has been successfully completed on multiple occasions. The second stage is the target zone, where a horizontally projected cyan LED beacon marks an array of 5 targets. The Prowler homes in on this beacon, and determines its distance from the beacon using its onboard video camera. Next, it releases two markers when it is over the targets. After the vehicle releases its two markers it makes a hard left towards the recovery zone, marked by a pinger. Using its passive acoustic system the Prowler homes in on the pinger. When the vehicle reaches the pinger it stops by reversing its drive motor and, being positively buoyant, rises to the surface of the recovery zone. A map of the arena for this mission is shown below in Figure 2. Figure 2 Competition arena map showing the practice and competition sides. The Team The team is comprised of students studying both Ocean and Electrical Engineering. The team works hard and tirelessly in the hopes of making a vehicle that will succeed at all of the tasks presented to it, as well as leave the next year's team with a strong base vehicle to work off of. One year of preparation and independent study has turned the Prowler V into a more powerful and efficient intelligent machine than its predecessors. Control System The control system for the Prowler V is comprised of sensors, sensor processing modules, a sensor distribution module and system control module. The AUV control system layout is shown below in Figure 3. Figure 3 AUV control system. System Control Module – Savage Innovations OOPic-R Microcontroller The AUV system control board for the Prowler V is a single OOPic-R microcontroller, as seen in Figure 4 (next page). This popular "object oriented" microcontroller provides a stable and powerful building block for the Prowler V system control module. The OOPic is programmable in three language styles similar to Basic, C or Java. The OOPic offers an object oriented environment, making programming simple. A unique feature of the OOPic is its "virtual circuits" which allow objects to pass information to each other in the background of the script (multitask). In the Prowler V, the OOPic microcontroller simply runs through a program script customized for the mission based on sensory data and its internal clock. The microcontroller controls two servos that actuate the rudder and dive planes, the propulsion speed and direction, as well as the marker dropping mechanism based upon the sensory data. (Savage Innovations, 2004) Figure 4 Savage Innovations OOPic-R microcontroller. Sensor Distribution Module – 4 Channel 8 bit Parallel Multiplexer To keep the sensors from overloading the system control module, the sensors go through "sensory processing and distribution" modules. The sensory distribution module is a 4 channel 8 bit multiplexer, controlled by the system control module. This allows the control system to select the specific sensor to be read. The distribution module, designed by our team, acts as a switch track so that all of the sensors are able to share the same 8 bit parallel lines to the system control module. Sensor Processing Modules – Savage Innovations OOPic-R Microcontroller The sensory processing module is composed of several microcontrollers which transform the raw sensor data into something simple and usable by the system control module. Two microcontrollers are used in the Prowler V for the sensor processing module. One OOPic-R microcontroller takes the serial sentence output by the HMR3000 compass (described below) and transforms the relevant information (heading) into a 9 bit parallel number for the sensory distribution module. A second OOPic microcontroller takes the serial sentence output by the CMUcam and (described below) and transforms the relevant information (beacon coordinates) into a 9 bit parallel number for the sensory distribution module. The passive acoustics module (described below) has a built-in microcontroller processing module that transforms the raw data from the transducers into a 6 bit parallel number for the sensory distribution module. Sensors Passive Acoustics The passive acoustic system module on the vehicle is similar to an ultra-short baseline (USBL) navigation system. USBL is a type of navigation system that uses a single transponder and acoustic array to give the vehicle an idea of its location. The directions of the incoming acoustic signals in the horizontal and vertical planes are determined by differences in arrival times at each of the receivers. USBL is typically used for AUV docking procedures where the vehicle tracks a sound source using a transducer array in the nose cone. Generally, a USBL receiver consists of four transducers set up in a cross pattern. The passive array designed for this system is a variation of the USBL phase array. This system aligns three transducers across the horizontal plane of the nose cone. The arrival time of the signal is compared between each pair of transducers. This method gives a general left-right direction of the transponder (sound source) by dividing it into four regions. Looking at the array from a top view, transducer C is set roughly 1-2cm in front of transducers A and B. The reason for this is that it is assumed the vehicle is in the far field of the transponder; therefore plane waves will be reaching the array. By offsetting transducer C from the line between transducers A and B, it is possible to discern the two additional center regions CA and CB (shown in Figure 6). Figure 6 Transponder regions. The closer to the center line transducer C is, the tighter the center regions are. The input signal is conditioned by filtering, thereby isolating the desired bandwidth. The signal is then amplified and compared to a reference value. When a signal is detected, a NE555 timer circuit is triggered sending its output line high signaling an arrival time on that transducer. The arrival times of the three pairs of signals are compared using three Digital Flip-Flops (DFF), such as the Texas Instruments SN74HC74. A DFF is a device which has two stable states. The DFF changes state by monitoring two inputs and changing state with respect to which signal was first received. The results of the three DFFs are fed into a 6-bit register on a Pic microcontroller. The transducers used are omni-directional spherical 2 cm diameter units that originally were used in Sonobuoys. Sensors Compass Honeywell HMR3000 A Honeywell HMR3000 (shown in Figure 8) was used as an electronic compass for the Prowler V. This device uses three magneto-resistive magnetic sensors to form an X, Y, Z magnetometer and a fluid filled two axis tilt sensor. The tilt sensor is used to calculate a correction factor to compensate for the tilt error in the heading data. The HMR3000 outputs a NMEA 0813 standard sentence over RS232. NMEA stands for National Marine Electronics Association, which has defined electrical interface and data protocol for communications between marine instruments. The 0813 designates their general sentence format. The sampling sequence of the sensors is controlled by an onboard microcontroller. The tilt compensation along with offset, gain error, and temperature effect corrections give the sensor an accuracy of better than 0.5 degree with a 0.1 degree resolution. (Honeywell, Inc., 2001) Figure 8 Honeywell HMR3000 electronic compass. The HMR3000 has a number of self calibrating features such as a hard iron compensation function which corrects for effects felt by nearby ferrous objects or electro magnetic fields. The device is also encased in a non-magnetic metallic enclosure with a DB-9 connector. The metallic enclosure allows for a simple and rugged way method of mounting. With the DB-9 connector, the HMR3000 is easily interfaced with any RS-232 compatible device. The RS-232 settings can be changed via a programming mode using a PC interface. (Honeywell, Inc., 2001) The onboard microprocessor reads the five sensor inputs and filters them using an IIR filter. The heading value is calculated from these values at a frequency of 13.75 Hz. The heading data, along with the pitch and roll values, are then combined into a serial sentence and sent out through the RS-232 port. In this system, the HMR3000 is connected directly to a microcontroller in the sensory processing array. Sensors Pressure – Motorola MPX4115 The Motorola MPX4115 pressure sensor is used in for depth sensing in the Prowler V. The sensor, shown in Figure 9, is a board-mountable single port pressure sensor. This device can measure over a pressure range of 2.9 to 36.0 PSI gauge/absolute. Some of the benefits of this sensor are that it produces linear results in 95% of its operating range with extremely low hysteresis. The extremely low hysteresis is due to the single crystal silicon that these sensors are produced from, which has superior elastic characteristics. The maximum depth at which this device can operate is approximately 15 m. With a maximum rating of 36.0 PSI, subtracting 14.7 PSI (1 atm) yields 21.3 PSI, which is equivalent to approximately 15 m of water. (Motorola, Inc., 2001) Figure 9 – MPX4115 pressure sensor. The sensor is directly ported to the environment using a length of Tygon tubing through the pressure hull. A small loop is added to the tubing to slightly increase its length. This prevents fluid from the environment from coming into contact with the sensor, which would cause it to malfunction. The fact that the tube is filled with air (a compressible fluid) and not oil (or another incompressible fluid) was taken into account. It was determined that the error due to the compressibility of the air over such a short distance is not a first order source of error. The MPX4115 operates over a range of 0.0 to 5.0 VDC. When using a 10-bit A/D converter, as in the Prowler V, the operating voltage range is divided into 1024 steps; each step representing approximately 0.005 V. In the Prowler 5 the system has a resolution of 0.1 feet. One of the analog input lines on the system control microcontroller module is used to record the pressure data. (Bashour, 2001) Sensors Imaging – Carnegie-Mellon CMUcam In order to track in on the target array marked by a LED beacon, the CMUcam image acquisition and processing unit was used. Originally developed at Carnegie-Mellon University, the camera is a low power CCD module with a programmed microprocessor which offers a wide range of functions for the quick analysis of image data. Among these functions, the camera offers the ability to track a color, or simply detect one. Along with a suitable 143x80 pixel resolution, the CMUcam provides the ideal simplicity that is essential for interfacing to the sensory microcontrollers. The CMUcam is shown below in Figure 10. Figure 0 CMUcam used to home in on the LED beacon. The Prowler V uses the CMUcam to track the color output by the LED beacon by using a subroutine within the CMUcam. The CMUcam outputs the pixel coordinates of the beacon and from this the vehicle determines which direction it must proceed to reach the beacon. The output of the CMUcam is in the form of a serial sentence. (Carnegie Mellon University, 2004) Electrical System The Prowler V operates on a rechargeable 14.7 Volt nickel metal hydride battery pack. It is rechargeable through the 6 pin wet plug located on the top of the vehicle. This makes it possible to recharge the vehicle without removing the pressure hull (although the nose cone is removed for off gassing safety reasons). The wet plug also makes it possible to run the entire vehicle off of external power for bench and tether testing. The microcontrollers and sensors run off 5 VDC which is created by a 5 Volt DC to DC converter. A DC to DC converter was used, rather than a linear voltage regulator, for increased efficiency and its built-in noise and spike reduction circuitry. To further increase efficiency, the motor runs off of the raw battery rather than another regulated supply. Mechanical System In designing the Prowler V, materials were chosen which ensured its strength to withstand pressure for the depth at which it is to operate, as well as make the vehicle watertight, so as not to damage vital electrical components within. Another major concern for the vehicle is corrosion, which can occur with certain materials or a combination of them in a water environment. Delrin® (Acetal plastic) was used in the construction of both nose and tail cone sections, and it was chosen for its strength, dimensional stability, moisture resistance, high modulus of elasticity, as well as excellent machinability. The push-rod system used to control the turning rudder behind the propeller was crafted using 316 steel. This is a choice made after last year’s experience, where aluminum was used and ultimately resulted in corrosion. The acoustic module on the nose cone was crafted out of Uralite®, which is a brand of urethane. This material has excellent acoustic properties, as well as enough strength to absorb a potential impact. The chassis of the vehicle is made completely of aluminum to reduce any magnetic effects that could interfere with the compass. The aluminum chassis also provides a lightweight, strong, and modifiable frame to mount the components and a common ground to decrease electrical noise in the system. The hydrodynamics of the Prowler are very important, and have been studied in detail over the past few years. Given the Prowler’s classification as a “flyer” type of AUV, it must be able to move quickly through the water, yet with fine control otherwise the mission is compromised. Good hydrodynamic design increases performance as well as efficiency. Every edition of the Prowler vehicle, since its inception in 2000, has had a long, cylindrical body tube capped with a rounded nose cone at the font and a streamlined tail cone. The tail cone design also allows for increased flow to the propeller, yielding more thrust per propeller rotation. This thrust is further amplified with the use of a kort nozzle. This nozzle acts as a shroud around the propeller, forcing water straight through the propeller, negating losses from water exiting perpendicular from the desired flow path. Conclusion As can be deduced, the Prowler V is a more efficient, reliable and predictable vehicle modeled off of the previous vehicle, Prowler IV. Through electrical , mechanical, and intelligence modifications the Prowler V will prove to be an impressive machine designed to satisfy the mission requirements as stated by AUVSI and ONR's Autonomous Underwater Vehicle Competition Rules.