Go-Go Gadget Tactical Robot Submarine
Christopher Neal1, Jacob Bock2, Jeremy Stark3
Abstract
A tactical remotely controlled surveillance submarine is a novel solution for jobs involving dangerous underwater conditions. Divers in aquatic conditions are subjected to conditions that could pose serious dangers; these risks to human life could be avoided completely by substituting a remotely controlled machine for video surveillance (other sensors which could be easily added through serial communication interfaces contained internally or positioned outside of the submarine). By using internal batteries to power the device and tethered communication to the surface of a water environment, video and other sensor data can be remotely sent and the device can be controlled by a remote computer, with the operator out of any danger posed by the aquatic conditions. By choosing readily available commercial products and common materials, such as PVC tubing for a submarine body, the machine can be kept at a low material cost and can achieve financial viability as replacement for people in dangerous conditions. Extensibility can be maintained through sensors that can be added through serial interfaces connecting into multiple microprocessors inside the submarine, so that in future designs or as needed by new applications, more observational possibilities can be added. Autonomous control can be developed in future models of the original design of the Go-Go Gadget submarine by implementing a sonar network and developing and troubleshooting an autonomous control algorithm for the submarine. The proof-of-concept design in the present paper can be remotely operated through a tethered control, can send and receive communications and directions, and can remotely deliver information about the submerged environment.
I. Introduction
This is a project to design a tactical submarine for multiple applications such as evaluating dams and ecological habitats, and surveying environments hazardous to humans. The submarine will incorporate features such as multiple operation modes, easy deployability, low cost development and production, and extensibility. Automated control will be determined using a network of sensors in conjunction with a series of software algorithms. The sensors will include sonar, GPS, accelerometers, temperature gauges, and others. The vehicle will feature a modular system to allow the powering of cameras and sensors. Additionally, the system will have a modular connection interface for additional sensors/accessories. A team of two should be able to deposit and remove the system into/out of a body of water.
II. Research
Much research was required before beginning the design process of the Go-Go Gadget Submarine. The research was broken into sections, in accordance with the expertise of the team members, to make the research easier. Mr. Stark, a student of electrical engineering, researched battery technology and power requirements, body materials, and propulsion; Mr. Neal and Mr. Bock, students of computer engineering, researched microcontrollers; and Mr. Neal also researched ballast tanks.
One of the best resources found was a paper written by Georgia Tech students covering their robotic submarine, The Mongoose. The Mongoose was Georgia Tech's entry into the 2005 AUVSI (Association for Unmanned Vehicle Systems International) and ONR Autonomous Underwater Vehicle Competition and was designed using inexpensive technology.1
A. Body materials
The simplest method for designing the hull of the submarine appears to be using PVC pipe. The Georgia Tech Marine Robotics Group’s Mongoose robotic submarine used a clear PVC pipe as the main body. This allowed for cameras mounted anywhere through the main body of the submarine, and visual inspection of the inner-workings. The size of the PVC body will be largely limited by the size of the batteries. The following tables have been compiled from data sheets given from the websites of Harvel PVC manufacturers.
B. Ballast tanks
Many submarines use multiple ballast tanks13 in order to distribute the changing weight throughout the body. While a large nuclear submarine may use about four tanks arranged along the bottom, the Go-Go Gadget Submarine can accomplish a similar goal with the use of two.
Pumps will fill or empty the tanks, which will control the buoyancy of the vessel. These pumps can be controlled using a pressure sensor and an Arduino microcontroller. Using pumps and ballast tanks, rather than propellers pushing downward, will reduce power consumption in the batteries in the long term and allow the submarine to use the energy in better places.
C.Propulsion
Submarine propulsion can be achieved through many different means, generally fitting to the use of the sub. The most commonly used form of propulsion is a propeller powered by a DC motor. For larger scale projects, or a need for faster speeds, pump jets are used. Some research and few projects have been related to magnetohydrodynamic (MHD) drives, which is a form of propulsion using electric and magnetic fields to provide movement.
DC motors can be implemented into a submarine drive system by attaching a propeller to the shaft of a DC motor and waterproofing the motor. A large amount of current will be needed for the motor to run, which will require a large battery capacity. Switching transistors and logic will have to be used to ensure proper stopping and reversing of the motors. Waterproofing a DC motor could prove to be troublesome because of special materials and product requirements. An alternative to constructing a waterproof DC motor propeller system was found in a design by the Georgia Tech Marine Robotics Group.1 This project used a Linux-based low-cost robotic submarine with commercial trolling motors used as propulsion. Using trolling motors would eliminate the need to waterproof the motors, allowing more time to focus on the design of the electrical system. A sufficiently large trolling motor can be purchased for around $150 through local sporting good stores.
The typical method of DC motor control involves the use of an H-bridge (Figure 1). An H-bridge is the link between digital circuitry and mechanical action and controls the rotational movement of a DC motor.
The term "H-bridge" is derived from the typical graphical representation of such a circuit. An H-bridge is built with four switches. When the top left and bottom right switches are closed (and the other two open) a positive voltage will be applied across the motor. By inverting all switches, this voltage is reversed, allowing reverse operation of the motor.
Pump jet and MHD propulsion, although both excellent ideas, require a large amount of testing and design; the electromechanical and fluid dynamics designs are simply out of the scope of this project. Pump jet propulsion requires a large amount of power and so becomes inefficient at low speeds. Another disadvantage is that clogging becomes a major factor due to the need for water intakes.
MHD propulsion uses the Lorentz force to create a magnetic field through two parallel conductors that propels water and causes the vessel to move.6 This design has no moving parts, but requires a large amount of power to generate the electric and magnetic fields needed to move the submarine at any feasible speed. The power requirements cause this method to be slower and require more navigational devices to be used for turning.
D.Microcontroller
The Georgia Tech Mongoose
The Internet is a great source of information concerning anything and everything robotics, including microcontrollers. One such resource is Arduino.cc, the support site for the Arduino development platform. The Arduino is an open-source platform with a relatively small learning curve. The Arduino uses a bootloader and a C derivative, eliminating much of the hassle that comes with some lower-level chips like the PIC or PICAXE. While the bootloader aims to ease the design process, the Arduino is quite powerful and not just for hobbyists. Enthusiasts and professionals alike have used the Arduino for home automation, robotics, motor control, etc. The Arduino features fourteen digital I/O pins,12 six analog input pins, and six analog output pins. The six analog output pins are actually part of the fourteen digital I/O pins, which can be used for Pulse Width Modulation (PWM).14 This uses a square waveform to vary its average value. PWM is commonly applied in order to regulate power delivery, allowing the load to have controlled power. In the case of the Go-Go Gadget Submarine, this will allow the motors to speed up and slow down. Using two boards will allow one board to control the motors and pumps. A second board will control all sensors and navigation, reducing the load on any one microprocessor. While a PIC or other low-level microcontroller may be the more common choice for professional robotics, the support for the Arduino in the form of tutorials, data sheets, and forums is both invaluable and unavailable for other microcontrollers. The low cost, versatility, relatively small learning curve, and great support for the Arduino made it the clear choice for the Go-Go Gadget Submarine's microcontroller.
The analog I/O will allow for reading and interpretation of sensors. Among the sensors to be implemented are multiple sonar sensors, an accelerometer, and GPS. Using the array of sonar sensors, the submarine will be able to detect obstacles and avoid them while on a pre-planned route. The GPS will allow the submarine to navigate from point to point on a map while above the water or at shallow depths.
The Arduino uses the C programming language, a standard among such devices. It is also capable of handling C++ with the use of various third-party compilers. Programming the I/O is relatively simple using this microcontroller, so most of our attention can be paid to collision detection and route planning.
Various internet forums proved extremely helpful in researching the digital electronics portion of the project. Two such sites are SparkFun.com and SocietyofRobots.com, both of which are aimed towards enthusiasts. SparkFun is a well-known supplier of electronics components including microcontrollers, sensors, soldering components, etc. In addition to fair prices on their products, SparkFun has an informative set of tutorials and a thriving online forum discussing everything from basic 'Hello World' introductory applications to advanced wireless functionality. The Society of Robots is an especially helpful resource for understanding the foundation of robotics including actuators, sensors, batteries, and microcontrollers. Society of Robots also provides valuable information for designing printed circuit boards for robotics applications. The site also has numerous calculators necessary for efficient robot applications. Similar to SparkFun, the Society of Robots has a thriving online forum.
D. Power Requirements
Power requirements for the submarine will be a major limiting factor in the design and scale of the project. In general a battery will be used to provide a DC power supply for a robotics project. The types of batteries used in other AUV projects are lead acid, lithium-ion, and lithium-ion polymer batteries. A new battery technology, lithium iron phosphate, will also be investigated. These battery types all vary with size, weight, power storage capacities, and price.
Lead acid batteries are by far the most common and inexpensive. These are commonly used as car batteries and are durable and long lasting. The energy to weight ration for lead acid batteries is approximately 30-40Wh/kg2. Although these batteries do have a relatively large energy density, because they are made of lead, the amount of weight added will constitute a large portion of the weight of the submarine if these are used.
Lithium-ion batteries are commonly used in mobile technologies like older cellular phones and laptop computers. These batteries have a higher energy to weight value, 160Wh/kg, and maintain a charge over extended periods of time.3 Lithium-ion batteries are more sensitive to changes in temperature, overcharging, and rapid discharge, and have been known to explode under these conditions. Using lithium-ion batteries will require many safety measures to assure proper handling of the charging, discharging, heat, and pressure obstacles that may come up in the design of the submarine.
Lithium-ion polymer batteries are a newer version of the lithium-ion batteries used in mobile technologies. These have an energy to weight rating of up to 200Wh/kg.4 Newer-model cellular devices that require longer-lasting batteries are using lithium-ion polymer batteries. The batteries have similar limitations as lithium-ion: fast charging and discharging can cause device failure or explosion. If this technology is used in the design, precautions will again need to be taken in the power system.
Lithium iron phosphate batteries are a new technology that is being used in limited applications. The energy to weight rating for these batteries is approximately 70-90Wh/kg.5 Although this rating is significantly lower than lithium-ion technologies, these batteries have a faster charge time and can tolerate extreme situations without the dangers of lithium-ion batteries. These batteries are being manufactured by several companies, such as Valence Energy Storage Solutions and A123 Systems, and are available through a small number of commercial outlets.
III. Goals and Objectives
The goal of our project is to produce a modular, tactical submarine for various applications such as surveying/evaluating dams and ecological habitats, and surveying environments hazardous to humans. The objective is to build a submarine which will incorporate features such as multiple operation modes, easy deployability, low cost development and production, and extensibility. The submarine system will provide autonomous and remote wireless and tethered control, as well as route planning control. The body of the submarine consists of large PVC tubing, measuring between 10” and 12” in diameter. Automated control will be determined using a network of sensors in conjunction with a series of software algorithms programmed in C/C++. The sensors will include sonar, GPS, accelerometers, temperature gauges, as well as others, all of which will be controlled with dual Arduino microcontrollers. Propulsion of the sub will be handled with three trolling motors. The trolling motors’ direction and speed will be controlled with an H-bridge. The vehicle will feature a modular system that will allow the powering of cameras and sensors. Lead acid and/or lithium-phosphate batteries will provide power to the motors and all electronics. Additionally, the system will have a modular connection interface, a modified USB connection, for additional sensors/accessories. A team of two will be able to deposit and remove the system into and out of a body of water by way of handles on the PVC body.
IV. Design Process
A. Chassis Design
A 10” PVC tube of approximately 5 feet long was chosen to use as the housing. This allows for a high level of pressure tolerance while providing a desirable amount of interior space. Two wooden rails were mounted within the pipe body. These rails allowed for a solid piece of 4 ½” pine to be moved inside. The rail system allows for the entire contents of the submarine to be moved in and out from one end, allowing the rest of the structure to be sealed, ensuring waterproofness.
Two solutions may be considered for mounting the motors to the chassis. One solution is to create a super-structure out of tubular metal (aluminum or steel) that the motors would attach to. Another solution is to attach the two motors through the inside of the tube. At present, the super-structure is the preferred solution as it allows us more room for the sliding shelf that holds all components. The super-structure would also allow for easier transport, as the motors would fold up.
B. Propulsion Design
The most common implementation of DC motor control in robotics is through the use of an H-Bridge. An H-bridge is the link between digital circuitry and mechanical action and controls the rotational movement of a DC motor.
While H-Bridges can easily be constructed using transistors and the above diagram, an integrated circuit h-bridge was chosen for a multitude of reasons. An integrated IC is far easier for prototyping purposes, has better heat dissipation, and has a much smaller footprint.
Samples of the LMD18200 H-Bridge (1) were ordered from National Semiconductor. The LMD18200 is rated at 3 Amps and 55 Volts, well above the Go-Go Gadget Submarine’s maximum operating environment. Some of the features of the LMD18200 include automatic thermal shutdown to prevent damage, PWM speed control, dual outputs, and braking.
Initial testing of the H-Bridge was done by modifying a simple LED fading program. The program outputs a PWM frequency starting at 0 Hz, then slowly increases to 500 Hz. Once reaching 500 Hz, the frequency slowly decreases to 0 Hz, then starts again.
H-bridge ICs were tested for motors, forward, left, right, and backwards movement through control by the microprocessor. Although only tested for a short period of time, considerable heat built up in the IC. Several options for heat transfer and removal will be researched for the best solution. Currently there are several heat sinks available that can be readily tested. Instead of delivering a PWM signal to control speed, the Arduino will simply output a digital high to turn the motors on full speed. Possible implementation of speed control is the use of a multiplexer to select which of the speed control wires receives voltage (allowing high, medium, and low velocities).
C. Power System Design
One trolling motor, Minn Kota Endura 30HC, was purchased, along with a small 8AH 12V SLA battery. This system has been tested in dry conditions to determine the power drain of the trolling motors over a long period of time.
The battery maintained a sufficient voltage over ninety minutes. This amount of time with a 7AH battery shows that a battery with a larger capacity will be sufficient for the goals outlined for the operating time of the submarine. Using Power Sonic PS-12180 SLA 18AH 12V batteries to power the motor allows the submarine to operate for up to an hour. The larger-capacity battery should also make up for any added force from waves in water.
The Arduino is to be powered by USB from the internal laptop computer. This will allow for better control of voltage changes to the sensitive microprocessor. No testing should need to be done for this power requirement as it has been previously shown that the Arduino can run from low power sources, like AA batteries, for long time periods. A battery charging system was purchased from Power Sonic, the battery manufacturer, to ensure safe and quick charging of the batteries.
The power supply system for the Go-Go Gadget Submarine will need to have two DC voltage levels, +5V for the Arduino and sensors, and +12V for the motors and bilge pumps. The power is being provided by 12V sealed lead acid batteries, hence a DC/DC converter will be used to convert 12V DC power to 5V DC for the Arduino and sensors.
The 5V power supply was tested with the H-bridge IC, using it for control as a high logic signal. Voltage was steady at 4.98V. When the final H-bridge ICs arrive, the final power supply system will be assembled and mounted to the body of the submarine with the H-bridge assembly.
Because the computer for control and communications is inside the submarine, the 5V power supply device is not needed. Arduino microprocessors and peripheral devices will be powered through USB from the internal submarine computer.
D. Digital Electronics
Continued sensor testing has allowed for a greater understanding of the available components and how they will work with the overall system. The Flexiforce pressure sensor has been a large task. The sensor, which can read up to 100 lbs, is intended to be placed on the wings of the submarine in order to indicate the depth. It is piezoresistive (1), which varies from infinite to about 300k Ohms. The Flexiforce (2) sensor may be abandoned in the future in order to use another sensor that Mr. Bock researched.
GPS functionality and temperature sensing are working as planned with both reporting data through serial to HyperTerminal. While the temperature sensors are able to report data at any time, and thus send data at any time and any speed, the GPS receives data intermittently and thus the baud rate must be set at a relatively slow 4800. An Arduino program that sends GPS data and temperature data was created to parse the NMEA sentence sent by the GPS and the analog data from the LM35 temperature sensor. While the GPS program outputs numerous fields such as distance to destination, speed, direction, time, and data, the GPS latitude and longitude coordinates are the only ones needed for the submarine.
While implementing the GPS and temperature sensors was relatively easy, it was not without complications. First, the LM35 temperature sensor requires a 5V supply while the GPS requires a 3.3V supply. Inputting 3.3V into the LM35 gives inaccurate readings, while a 5V supply into the GPS can destroy the electronics. Another obstacle was the surface mount wire connector on the GPS. An internet search into a breakout cable for the diminutive pins provided zero results. While using a very high gauge wire works for breadboarding purposes, a more permanent solution must be acquired. Operator error was also an issue. The initial GPS program was thought incorrect as the GPS never received a lock (acquiring at least 4 GPS satellites). In reality, the GPS was connecting to satellites, but the Time to First Lock (TTFL) of the GPS is nearly 3 minutes.
The speed of the Arduino proved to be sufficient for the sensor network with a 16 MHz clock oscillation. The memory, on the other hand, was upgraded from the Atmega 168 to the Atmega 328, which features an increase in memory from 14K to 30K.
E. Sensor Design
Most commercially available Sonar systems are for large marine systems, or smaller land based applications. Neither of these readily available systems will be suitable to add to the submarine without major modifications. A small underwater ultrasonic transmitter and receiver combination has been used in similar applications.1 Since this system is only an underwater transceiver, it will require more work in programming the ranging logic and implementing this into the movement algorithm of the submarine.
After researching possible alternatives to buying expensive commercial underwater ultrasonic transducers and hydrophones, another possibility was found. A hydrophone was made from a simple set of ear-bud headphones by using silicone sealant to insulate and waterproof the device.1 This method could be applied to cheap ultrasonic transducer receiver kits that are used in ultrasonic ranging methods. An ideal device is the ultrasonic transducer kit, from Prowave, an ultrasonic transducer manufacturer.2 These transducers work in the 40 kHz range, and could be appropriately and easily adapted to underwater Sonar applications, if waterproofed. This would also require a driver circuit to power and amplify the signal being produced by the ultrasonic transducers, but most products would require this also. More testing will need to be done this week to show that this method of waterproofing will work for the application.
A set of 40 kHz ultrasonic transducer/receivers has been procured and ran through
preliminary testing for adaptation for underwater Sonar use. The transmitter and receiver were first tested in a system for in air transmission at varying distances and were found to work at distances greater than 3 m with a strong unfiltered reception signal using a standard 40 kHz signal generator. After this test was performed, silicone sealant was placed over the transmitter and receiver for waterproofing and the procedure was repeated. Although the signal was greatly degraded due to noise from the uneven application of silicone, the frequency signal was still discernable. The test was then recreated in a large beaker to test for water transmission and the results were similar. Through this short proof of concept, a 40 kHz driver and 40 kHz filtering system can be developed and possibly adapted into a sonar navigation system for the Go-Go-Gadget submarine in the next phase of development.
Two Microsoft Lifecams were purchased and tested. Initially it was thought that a .NET C# application could be written to control the cameras with little effort. However, the WIA library (the webcam library provided by Microsoft) is only capable of capturing sub-VGA resolution (640x480 pixels) images. To avoid the low-resolution barrier, the proprietary webcam software provided by Microsoft will be used to display and capture images and videos. A .NET application will be created that docks the viewing window with another window displaying information such as temperature, depth, etc.
One of the Lifecams was hacked into an infrared camera by simply removing the IR filter. The camera was disassembled and the lens removed with little effort.
Sensors used for navigation such as GPS and a digital compass were implemented in order to provide a sense of direction and location. Components were chosen specifically for their ability to communicate serially, as is a requirement of the Arduino.
Two temperature sensors were implemented in order to measure both the internal temperature near the H-bridge circuits and the water temperature just above the submarine's body.
Wireless communication was not implemented in this design due to the constraints of wireless communications in an underwater environment. Sufficient bandwidth would prove difficult for proper signaling of commands. This can be further researched and improved upon in later models.
V. Build
A. Build Chassis
Because the PVC tube required no construction, the group was allowed to pay more attention to the superstructure used to support the motors. This was constructed using aluminum tubing and bolts over the course of several weeks.
A chassis has been constructed using one of two 5-foot PVC pipes that were ordered earlier during the design phase. Inside the PVC pipe was mounted a wooden board to mount the batteries, the laptop, and any other components that need to be accessed. This wooden board is designed to slide in and out of the submarine’s opening, allowing the parts located deep within the interior to be accessed.
Schedule 40 PVC end caps have been selected for the sealed end of the submarine. This will be permanently fixed to and end of the PVC structure. To allow for video, a window has been placed in the end cap. The window was cut from Lexan, and fixed to the PVC with Gorilla Glue. The window will be constructed of two Lexan pieces, one mounted outside of the hole drilled into the end cap, and one inside. After drying the end cap will be fixed to the submarine permanently.
An expansion plug has been selected for the accessible end of the submarine. The plug was ordered and has arrived, but no testing was done other than inspecting the plug for proper size and operation.
A superstructure was constructed around the body to support the weight of the motors, and provide portability by allowing the motors to fold in. The superstructure was made from 1 inch x 1 inch aluminum tubing and bolted to the body to form the rigid structure. Aluminum was chosen for its easy workability, strength, and relatively light weight. All drill holes will be sealed to ensure that the submarine does not become an anchor. Gorilla Glue, a widely available all-purpose sealant/glue, will be used as the primary sealant for its hardness and water-sealing capabilities.
The superstructure was constructed with the ability to ‘swing’ the forward-facing motors upward for portability. With the arms folded upwards, the width of the submarine is such that it can be transported in a standard pickup truck (48 inches).
Along with the forward-facing motors, the vertical movement motor will be mounted to the superstructure. The vertical movement motor will be attached to the underside of the sub with a removable ‘cage’ that allows for easier portability. When the cage is detached from the superstructure, it will still be tethered by the motor control wires.
B. Assemble Motors
A solution for the ascension and descension of the submarine was reached by using a motor to control the up and down movement of the submarine after achieving neutral buoyancy of the sub. A design was implemented and constructed by attaching a motor to the superstructure supporting the other motors. This motor was mounted using the same aluminum tubing, and will be controlled using a separate motor controller circuit that is operated by the microprocessor.
The motors have been mounted to the superstructure successfully and will still be allowed freedom to swing out of place for better transportability when finally placed. The motors were permanently affixed to the submarine chassis as soon as the submarine was sealed. The motor shafts were filled with expanding pipe foam, then glued to avoid any water damage within the motor. The wires of the motors were fed through .5 in holes, which were also sealed with glue.
C. Assemble Electronics
A 5V power supply was assembled this week using a DC/DC converter (V7805-500, V-Infinity). Due to the low amount of power loss in this converter, there is no need for a heat sink. The power supply uses one 12V source and has outputs for two 5V, 500 mA connections. The supply was tested for tolerances and spikes.
Due to the large amount of heat produced by the H-bridge IC, heat sinks were needed to dissipate heat generated by these chips. After running these while testing the motors for a brief time, the IC quickly reached shut-off temperatures listed in the data sheet. To resolve this problem, the H-bridge circuits were mounted to large heat sinks salvaged from an old RF amplifier. After mounting the IC to the heat sink, it was tested and ran at safe operating temperatures for extended periods of time. The current through the IC ran at around a maximum of 3A at full speed through the testing.
H-bridge motor control circuits were completed for all motors. Heat sinks were assembled and circuits have been tested for proper function and temperature control. Motors have been mounted to the superstructure of the submarine. Motor control circuits are mounted internally.
Assembly of the sensor network was relatively simple due to the fact that all circuits were built and tested during the development phase. Mr. Bock was required to create a small board to house the GPS and compass, which allowed all the sensors to be held firmly to the wooden board.
As power requirements should be no problem with the excess battery capacity provided by the current batteries and the battery life of the laptop, a 100 foot Ethernet cable was used as a tether and remote connection to the submarine control. The 100 foot crossover cable was used to form a network connection between the remote controlling computer and the internal submarine computer. A VNC connection was used to remotely log into the internal submarine computer and control devices and motors.
D.Software Implementation
A C# .NET front-end was built to access information sent from the Arduino’s sensors. A simple windows form was created to display information such as GPS, compass direction, temperature, and depth. The program also starts the Microsoft LifeCam software and closes it upon exiting the program.
The program uses the Microsoft Serial.IO library provided with .NET. Implementation is simple as opening a serial port requires only the port name, baud rate, parity information, and stopping bit information. More data than necessary is sent from the GPS. The needed data is simply parsed from the supplied data and fed into the windows form. An image of the program is provided below. Additional functionality of the program will include tethered control of the submarine.
Implementation of the compass module, HMC6352, required extra research into Arduino libraries. The compass module connects with the I2C interface on most electronics. The Arduino uses the I2C interface with the use of the Wire libraries (available in the default Arduino IDE). Online research turned up various tutorials on how to implement the compass. With little editing, the code from the tutorials worked well. The module connects through the Wire/I2C interface with 2 data lines, 1 ground line, and 1 3.3 Vcc line. The module outputs a floating point value in degrees relative to true north (ie 0 is north, 90 is east, and so on).
Once the compass was implemented with the Arduino, the HUD application was expanded to provide the information to the user. Using a similar method as the existing GPS and temperature implementations, the compass information was exhibited in the HUD. Along with showing the degree value, the shown data is appended with a relative direction based on the degree value (ie 0>= degree >90 is NE).
The control interface is an addition to the HUD that will allow the user to control the submarine's movement using either keyboard commands or mouse clicks. This sends serial data to the microcontroller. The Arduino will then decode the message and set a value to one of its pins that correspond with the Direction, Brake, and PWM pins on the H-bridges. This is done using the analogWrite(); function built into the Arduino's library. The control interface will send values in the format "F", "B", "L", "R", "U", "D", and "S" to activate the pins on each H-bridge to send the motor signals to move forward, backward, left, right, up, down, and stop, respectively. For example, if a signal "B" is sent to the microcontroller, it will determine this is the command to move back and set the values of Direction, Brake, and PWM for the two side motors, also making sure that the other signals are not enabled at the same time. Once the signal is received and the value is set, it will remain at that value unless otherwise signaled. The motor control portion of the HUD and the Arduino code are finished and ready to be tested. The code from the compass module has been compiled into the HUD code with the motor controls.
VI. Testing
The submarine was tested for leaks in a pool. To avoid any possible damage to electronics, the submarine was tested without any internal gadgetry included including the laptop, batteries, and Arduinos. To test the submarine, Kimwipes (a tissue-like paper used in Chemistry) were placed inside the submarine and the expanding plug was fitted into the open end and the submarine was dropped into the pool. Due to the lack of batteries (~50 lbs) and the large displacement of the submarine, Mr. Neal and Mr. Bock had to push the submarine deeper to counteract the ~200 lbs of buoyancy.
Initially the submarine was held under water for 3 minutes without any movement. The sub was then removed, the plug removed, and the tissue inspected. No leaks were observed. Next, the submarine was placed back in the water and moved around vigorously to simulate movement through the water. The submarine was rolled frequently. This testing continued for approximately 15 minutes and again the submarine was removed and the Kimwipes inspected. Again, no leaks were observed. Further testing is required to tweak the submarine's weight for neutral buoyancy. Once neutral buoyancy is met, ascension and diving speeds can be tested.
The submarine was able to be controlled remotely as specified in the original goals. After trouble-shooting the device, through the use of a remote laptop computer tethered with the submarine, controls for motion in water were realized successfully, and recorded on video. The data from sensors was also successfully viewed through the display software developed.
Power requirements were met and exceeded through the use of larger batteries. After testing and troubleshooting the submarine all afternoon in the Aquatics Center, battery levels were still acceptable for more use.
Sensors proved to be accurate with respect to the environment in which they were tested and were therefore determined to be successful.
VII. Discussion
As a whole, it is believed that the design went well. Unfortunately, there were several major setbacks that caused the design to be modified. Because of this, there exist numerous future enhancements that would be desired if the project were to continue.
The absence of automation was a product of time constraint, but would be a desirable future enhancement. In the event that this was the case, engineers could add the autonomous navigation algorithms as well as the additional sensors necessary. This would include following preplanned routes and collision detection using sonar. A modification of this nature would require a large time contribution in order to develop and tweak the algorithms based on the environment in which it was tested. Therefore, this type of enhancement might not be feasible unless it undertaken as a project of its own.
Proper neutral buoyancy is another aspect that the group was unable to successfully implement. Due to problems with the internal components shifting, the leak made it unable for the team members to continue with the calibrations necessary to make the submarine neutrally buoyant. A change of this nature would not be difficult, but it would require some fine-tuning in order to get things correct.
Additionally, the modular connectivity would be another enhancement in the future. In order to implement this, a rail could be constructed using similar aluminum tubing which would allow other items such as external cameras to be mounted. Power could be provided using a connection to a fourth battery that could fit on the underside of the rail system near the other three batteries. Any necessary processing would be allowed through the Acer Aspire One mounted inside the submarine. A standardized connector could be developed in order to allow the devices to communicate and receive power.
VIII. Conclusion
The design and construction of a remotely controlled tactical submarine was realized in the form of the Go-Go Gadget submarine. The design was kept to a low cost, basic construction for functionality and expandability in future models. A large PVC pipe was used for a body with a rigid superstructure built around the body as a support for the motors, which were commercial trolling motors. The pipe was sealed at one end with a PVC endcap with a Lexan viewing window and an expanding plug was used to secure the end and allow for easy removal of internal components. Control was realized through the use of Arduino PIC microprocessors with various sensors and an h-bridge IC, to control motor speed and direction. Through the use of sealed lead acid batteries provided significant battery life for underwater operation and proved to be a cost-effective solution. The first design and proof-of-concept testing of the submarine accomplished these goals. Further designs can implement autonomous control with sonar and control programming or possible wireless communication through radio frequency transmissions, improving on the original proof of concept design.
IX. References
1. Georgia Tech Marine Robotics Group. ”Introducing the Mongoose -- a Linux-powered Submarine.” 2005.
2. Battery Datasheets. Power Sonic Company. 2008.
3. “Lithium-ion battery.” Wikipedia, the free encyclopedia. 2008.
4. “Lithium-ion polymer battery.” Wikipedia, the free encyclopedia. 2008.
5. U-Charge XP Data Sheet, Valence Energy Storage Solutions, 2008.
6. Petrik, M. “Results from a large scale MHD propulsion experiment.” Presented at the 30th Symposium on Engineering Aspects of Magnetohydrodynamics. Baltimore, MD. 29 Jun. - 2 Jul. 1992.
7. Stojanovic, Milica. “Underwater Wireless Communications: Current Achievements and
Research Challenges.” Massachusetts Institute of Technology, 2006. IEEE.
8. Pierce, A.D. “Acoustics: An Introduction to its Physical Principles and Applications.”
American Institute of Physics. New York. 1989.
9. “Underwater Acoustic Modem.” LinkQuest Inc. Advertisement. November 29, 2008.
10. Jaber, B. N. and Whitten T. J. “Recoverable Tethered Optical Fibre Submarine Communications.” Ultra Electronics Limited 2006. November 29, 2008
11. Keener-Chavis, Paula. “Expedition Purpose.” National Oceanic and Atmospheric Administration. June 12, 2008.
< http://oceanexplorer.noaa.gov/explorations/08auvfest/background/edu/purpose.html>
12. Banzi, Massimo. “Getting Started with Arduino.” OReilly Printing Press. October 2008.
13. “Ballast Tanks.” Wikipedia, the free encyclopedia. 2008.
14. “Pulse Width Modulation.” Wikipedia, the free encyclopedia. 2008.
1. Christopher Neal
Christopher Neal has a B.S. in Systems Engineering from the University of Arkansas Little Rock. Specializing in computer systems, he is proficient in multiple programming languages including Assembly, C/C++, C#, VB 6, VB.NET, Java, and SQL. His experience with microprocessors ranges from commercial Texas Instruments hardware to commonly available open source hardware.
2. Jacob Bock
Jacob Bock has a B.S. in Systems Engineering from the University of Arkansas Little Rock specializing in computer systems. Jacob is experienced in many programming languages including: C/C++, C#, Assembly, BASIC, Java, Magik, and VHDL/Verilog. He has experience with many microprocessors including Texas Instruments Digital Signal Processors, PIC, and PICAXE. He has taken many electronics courses including Circuits and Systems, Analog and Digital Circuits, Digital Systems, and Advanced Digital Systems.
3. Jeremy Stark
Jeremy Stark has B.S. in Systems Engineering and Physics from the University of Arkansas at Little Rock. Mr. Stark has been working in the Particulate Materials Science and Engineering Research group at the University of Arkansas at Little Rock in various research areas such as aerosol particle measurement instrumentation and signal processing, electrodynamic dust mitigation techniques, and other areas of energy research. He is currently working towards a Ph.D. in Electrical Engineering at Boston University.
This is epic work. Love the
This is epic work. Love the site too! Keep it up!