Autonomous Robotic Fish to detect Harmful Algae Blooms (HABS)

FINAL PROPOSAL:

Design Team:4 Team Facilitator: Dean Aslam

Team members: Taha Tareen, Jamie Jacobs, Stephen Garrett, Woodard Williams Eric Jackson, Carl Coppola, Robert Morris, Allen Eyler

Date: 02/4/09

EXECUTIVE SUMMARY

Aquatic ecosystems are undergoing dramatic changes due to human activities and change in climate, which results in environmental pollution and affects human wellbeing. The proliferation of Harmful Algal Blooms (HABs) is caused by cyanobacteria producing toxins which accumulate rapidly in water bodies, thus proposing a great danger to our lives. At MichiganStateUniversity, the goal is to advance knowledge and transform lives. Through detecting harmful algal blooms (HABS) in three-dimensional water bodies, this project gives students an opportunity to achieve the goals of MSU by making the world a healthier place.

To achieve the goals set by the team, the most critical issues surrounding the project are the implementation of Graphical User Interface (GUI), digital signal controller (DSC), and the wireless integration. The following steps are to be taken as the basic requirements of the project. For bacteria detection the HAB sensor will be interfaced in to the circuit and the GUI will be updated accordingly. An infrared sensor (IR) will also be implemented for avoiding collisions in water bodies. The electrical casing inside the fish will be made for locking the circuit inside and is according to the conceptual design of the robotic fish itself which also has a compass integrated for the direction movement. Finally demonstrating the fish in the water tank or a lake will provide the results and ensure success.

The final design is a smooth body fish, similar to the current shape, with two fins used for balance and one bottom fin used as a rudder. The bottom rudder fin will be used for direction control. Whiskers will be placed on the front of the fish for collision detection. The PCB, battery pack, and servo will be contained in an aluminum water tight container inside the fish body. The bulk of the weight will be distributed towards the middle of the fish to provide the best weight distribution scheme. Upon completion the robotic fish will sense HAB, avoid harmful collisions, and swim in the intended direction.

TABLE OF CONTENTS
Introduction………………………………………………………………....2
Background ……………………………………………………………...... 2
Design Specifications…………………………………………………...... 2-6
Table 1 and 2…………………………………………………...... 3
Table 3 and 4…………………………………………………………...4
Table 5,6 and 7…………………………………………………………5
FAST Diagram……………………………………………………………...6-7
Figure 1………………………………………………………………….7
Conceptual Designs……………………………………………………...... 7-10
Figure 2 and 3…………………………………………………………..8
Figure 4 and 5…………………………………………………………..9
Figure 6 and 7…………………………………………………………..10
Rankings of Conceptual Design………………………………………...... 11
Table 8 and 9…………………………………………………………..11
Proposed Design Solution…………………………………………………..12
Figure 8………………………………………………………………...12
Risk Analysis……………………………………………………………….12-13
Project Management Plan…………………………………………………..13-15
Budget………………………………………………………………………..15
Table 10…………………………………………………………………15

INTRODUCTION

With the environment constantly changing due to human abuse of the planet, scientists are constantly faced with the challenge of coming up with new, more effective methods of understanding/predicting an ecosystem’s response time to global change. One of the more severe concerns of the planet is the aquatic ecosystems. As a result of different contaminants and toxins already disturbing our water, proper functionality of ecosystems and human welfare are dangerously at risk. More specifically, the abundance of harmful algal blooms (HABs) is becoming a more critical issue. In freshwater, HABs are caused by cyanobacteria producing potent toxins. These toxins can negatively impact water supplies and accumulate in fish. In the purpose of this project is to continue the development of an autonomous robotic fish that can detect harmful blooms. Consequently, this specific research should be able to open the door for better ways of monitoring the various water bodies of the Earth, and in the future, better ways of understanding ecosystem behavior.

BACKGROUND

In 2005, the robotic fish project was initiated by the Smart Microsystems Laboratory (SML) on the campus of MichiganStateUniversity. The mission of SML is to enable smarter, smaller, integrated systems by merging advanced modeling, control and design methodologies with novel materials and fabrication processes. The purpose of this task was to build small mobile platforms to be used in aquatic wireless sensor networks. Since August of 2005, SML has developed three generations of robotic fish.

The first design, G1, was developed by a senior design team. It was equipped with a microcontroller and wireless communication and was controlled through graphical user interface. The outer shell was a commercially available toy fish, modified to accommodate the circuitry. In August 2006, the G1 greatly improved on the specific problems such as space optimization and waterproofing of the circuit. The second version of the G1 was smaller size and weight. Additionally, it was equipped with a temperature sensor which allowed for a more life like, true mobile sensing capabilities. However, like the first model, the circuit was still confined to a toy fish shell. In 2007, a new prototype was introduced with ranging abilities and a custom built outer shell. New methods were also used to for better waterproofing the circuit. In 2008, the robotic fish was upgraded once again. This time the main focus of the upgrade consisted of computation capabilities. Instead of using a microcontroller, a digital signal controller was used which allowed for the implementation of two more complex ranging algorithms. Also, this design included an onboard battery power source which allowed for increased hours of run time. Throughout each generation of the robotic fish, the most significant prototype constraints included: the design of the outer shell to keep the circuit completely dry, mobility, and using the proper applications to transform the robotic fish from a mere fixed sensing circuit to a mobile robot.

DESIGN SPECIFICATIONS

The objective of this project is to build an autonomous robotic fish that is capable of three dimensional diving, wireless feedback controls, and detecting harmful algae blooms in diverse aquatic ecosystems. The product will exist as a prototype for developing a group of sensor carrying robotic fish that will monitor lakes and possibly help prevent deteriorating water quality. In order to accomplish the goal of our project the group determined which requested design criteria was feasible for the given time constraints. Table 1, shown below reflects these decisions.

Table 1. Feasible Design Criteria

To fulfill the objective the following constraints, ordered in terms of importance to the customer, must be satisfied:

1.)The robotic fish must be interfaced with a small sensor that can detect harmful algae blooms in water bodies.

The desirability of this aspect of the design is very high. The sensor detects cyanobacterial (algal) pigments that emit fluorescence when excited by certain light waves. The sensor would then record the concentration of algae and send information remotely to a laptop on shore.

HAB Sensors – Shown in Table 2 below

Cyclops 7: The Cyclops 7 is a submersible fluorometer designed for integration into a third party platform that supplies power and data logging. This product boasts low power consumption (less than 300mW), small size (weight: 5 oz; length: 4.3’’ x 0.9’’), and light sensitivity (dynamic range: low 0-5 µg/l; high 0-500 µg/l). The Cyclops 7 is designed to integrate with the C6 Multi-sensor platform. The sensor platform adds an extra 6 lbs and 10.2‘’ to the product size.

Phytoflash: The phytoflash is a submersible active fluorometer that can be used to detect natural concentrations of cyanobacteria in diverse water systems. The phytoflash is small in size (weight in water: 1.01 lbs; length: 12’’ x 3 ‘’), light sensitive (dynamic range: low 0-5 µg/l; high 0-100 µg/l), and has low power consumption (consumes less than 1W). The device does not require integration into a specific third party source.

Table 2. Feasible Sensors

2.)The graphical user interface (GUI), digital signal controller, and circuit board must be updated.

Currently the GUI, digital signal controller, and circuit board are designed for a fish that moves in two dimensions. For the robotic fish to work correctly, all components of the phase one, two dimensional fish must be upgraded and interfaced with sensors.

3.)The robotic fish should be equipped with wireless feedback controls.

The feedback control will allow the team to precisely know which direction the robotic fish is heading. This will be accomplished by interfacing the fish with a compass or magnetometer.

Direction Control

Servo: A servo is a small device that has an output shaft used to control the direction of the fish. The output shaft can be set to specific angular positions by sending a signal to the servo. As the signal changes the angular position of shaft also changes.

DC Motor: A small DC motor use output torque converted from a power source to direct the robotic fish. This option would be the most power consuming.

IPMC: Ionic polymer metal composite (IPMC) materials are highly active actuators that bend in the presence of low voltage. When the polarity of the voltage is reversed the polymer bends in the other direction. This bending motion resembles the movement of a fish and provides similar mobility in the water.

Table 3. Feasible Direction Control

4.)The robotic fish must be interfaced with a sensor that can detect approaching objects.

This sensor will protect the fish from crashing into rigid objects and becoming damaged. The sensor will detect an object and an event will occur in the circuitry that will allow the fish to turn and avoid the obstacle.

Collision Sensors- Shown in Table 4 below

IR Sensors:Collision avoidance sensors send infrared (IR) signals in the form of radio waves to detect an approaching object. The sensor then sends a signal back to the microcontroller allowing time for evasive action.

Whiskers: Artificial whiskers will be attached to the front of the robotic fish for navigation purposes. When the whiskers collide with an object the amount of deformation of the whisker shape will determine the voltage signal sent back to the microcontroller. This allows the robotic fish to either swim through trivial obstructions or avoid more stable structures.

Table 4. Feasible Collision Detection Sensors

5.)The robotic fish should possess a versatile packaging scheme and a suitable body shape.

With the help of team members with a mechanical engineering background, an upgraded robotic fish body and packaging can possibly enhance the state of the art of robotic fish. The robotic fish should resemble the appearance of an actual fish and the encasing must be water proof so that the electrical components will be protected.

Body Shape- Shown in Table 5 below

Jellyfish: The body of design one is shaped like a jellyfish with a long fin connected to the bottom. The design would float on the water and survey the surface. The fin would be there for stability in the water and direction control.

Tuna Fish: This body shape is composed of two dorsal fins for direction and a causal fin (tail) for propulsion. This body shape allows for hydrodynamic mobility in the water. The polymer in the tail would produce the propulsion for this design.

Manta Ray: The manta’s diamond shaped body is a perfect model for lake exploration. The pectoral fin would be used to guide the fish and keep it stable while a causal fin would propel the device in the water.

Tuna Fish 2: The same body shape as tuna fish 1 with no bottom rudder fin.

Table 5. Feasible Body Designs (pictures above)

Water Tight Compartment (WTC) Packaging – Shown in table 6 below

Pill Bottle: The shape of a pill bottle would keep out water with a twist on cap and house the electronic components inside. There might be extra unused room because of the width of a pill bottle.

Pop Bottle: The length and width of a pop bottle would be perfect for the tuna fish design. The only problem that would arise is cutting opening the bottle to store the components inside and then sealing it.

Pop Can: The pop can is made of aluminum so opening and soldering a part of

the can close would be an advantage in this design.

Table 6. Feasible Water Tight Compartment Designs

6.)The robotic fish should be capable of swimming 1.5 cm/s or faster.

To effectively monitor a lake within the battery life of the robot, the fish should be able to swim 1.5 cm/s. An increased speed will also enable the sensor to detect more harmful algae.

Propulsion – Shown in Table 7 below

Single IPMC: A single IPMC would be used like the causal fin of a fish. The horizontal bending motion of the material would propel the robotic fish forward.

Multiple IPMC: Multiple IPMC would provide more power to propulsion and allow the device to move faster.

Propellers: Propellers would be created using a dc motor. The torque outputted from the motor would rotate, displacing the water and moving the fish forward. The concern of this option is power consumption.

Table 7. Feasible Propulsion Designs

7.)The team must demonstrate the operation of the mobile sensor platform (including wireless communication and GUI) in a water tank.

Demonstration of the robotic fish in a tank will prove the capability of the project to move to an actual lake, if environmental conditions allow. Also, the operation of the fish in water will be tested.

The autonomous robotic fish will be designed based on the preceding criteria. To determine the desirability of a design, the following set of criteria has been developed:

1.)Function

This parameter determines if the design follows the mandatory constraints identified above. It is the first and most important aspect of the design.

2.)Size

The robotic fish body must be large enough to hold the HAB sensor and the other electrical components. Also, what must be considered is that if the fish is large in size and light in weight it will displace more water and have a difficult time submerging.

3.)Weight

The weight of the robotic fish design can affect ascension or descension in the water. To descend the fish must weigh more than the water it is displacing. To ascend the fish must be more buoyant than the water. A design with the ability to control weight while in the water is preferable.

4.)Energy Consumption

The robotic fish will run on batteries, so monitoring energy consumption is relevant to a good design. The rate of power consumed directly determines how long the device can stay in the water and detect harmful algae. A robotic fish that can use less energy and still be effective in its operation is highly desired.

5.)Reliability

This parameter concerns the ability of the design to perform tasks and operate in a consistent manner.

6.)Aesthetics

Appearance can have a large impact on marketability. A robotic fish that has a very close resemblance to a real fish can draw a lot of attention by looks alone.

7.)Delivery Date

A design that is too complicated may not be achievable by the delivery date. Each design must therefore be rated on its feasibility of completion time.

FAST DIAGRAM

In Figure 1 below is the FAST diagram for the design of the HAB sensor. A FAST diagram was not completed for the entire scope of the project because it would be extremely large.

Figure 1. Fast Diagram

CONCEPTUAL DESIGN DESCRIPTIONS

Using the preliminary matrices shown in the design specifications portion of this document four designs were proposed as solutions. The first proposed solution will be referred to as design 2a for the remainder of this paper. All four of our designs have the exact same placement of the HAB sensor and whiskers, which can be seen in figure 3 below. Design 2a has the body type as shown in Figure 2 below. The tail of the fish will consist of two IPMC (artificial muscle polymer). The two IMPCs will work in a “V” to “I” motion. In this way, IMPCs could provide more forward propulsion. The bottom fin will be used as a rudder, which is connected to a servo. This rudder will be used for direction control. The two fins coming out of the fish at angles will provide extra balance to the fish. Inside the fish body will be a controller PCB, a servo, and a battery pack. In design 2a, taking a birds eye view of the fish, the battery pack, servo, and PCB will be distributed from one side of the fish to the other as shown in Figure 3 below.

The second proposed design will be referred to as 2b for the remainder of the paper. The body shape is exactly the same as the body shape for design 2a. The main differences between the two designs are the placement of the battery pack, PCB, and servo, as well as the number of IPMCs. In design 2b there will only one IPMC providing forward propulsion. In this design the PCB will be placed above the servo and the batteries. The servo will be positioned in the middle of the fish between 2 separated batteries, see Figure 4 below for a sketch.