Project Name: RoboFish Charging Station (RCS) Project Number: P17250 Project Family: P16029, P16229, P15029, P14029 Start Term: 2161 End Term: 2165 Team Members Jack Moore - Mechanical Engineering - Project Manager Johan Nasution - Electrical Engineering - Control Systems Lead Brittany McCord - Electrical Engineering - Digital Communications Lead Garrett Burgwardt - Electrical Engineering - Electrical Engineering Lead Lu Ming Aung - Electrical Engineering - Software Lead Christopher Parker - Mechanical Engineering - Mechanical Lead Abstract The objective of this project has been to design a process that enables the biomimetic robotic fish (RoboFish) to engage in lengthier, time-sustainable missions. This will be accomplished through the use of a power-generating, floating docking station that is capable of charging the RoboFish called the RoboFish Charging Station (RCS). The fish must be able to first detect battery life capacity at a reasonable threshold, navigate flawlessly to the station, dock and charge batteries up to full capacity, and undock successfully to resume operation. Thus far, the final detailed design has been chosen through the use of Pugh Charts to minimize costs and complexity, as well as maximize aforementioned functionalities. Energy will be harvested using two solar panels, which will be enough to collect solar energy to recharge the 10Ah lead-acid battery established on the RCS. The charging process is completed through direct connection below the RCS, but above the water to minimize short-circuiting the cables. The RCS floats on the water through four dock floats on each corner of the square base, and include heavy weights as a means of anchor. Automation of the connection is processed through a small Arduino Pro. Finally, the complete process of the attachment includes the RoboFish swimming to a pole connected to the center of the RCS, grabs it to ensure correct location, releases pole and simultaneously floats until the charging port is guided by one of four funnels into the connection port to begin charging. Once the charging is complete, the RoboFish disconnects from the RCS, sinks back into the water, and continues its surveillance operation. Introduction An autonomous robotic fish (which will be referenced henceforth as RoboFish ) has been developed that employs hydraulic McKibben artificial muscles to power through water. The fish has been designed to minimize any mechanical or robotic appearance, and to be used in a multitude of applications, ranging from military surveillance to marine research. As an introduction to these operations, a sensor has been installed on the RoboFish - capable of detecting falling objects in the water. Furthermore, advanced mechanical talons were
established in order to intercept these objects before they reach the ground. However, the batteries that currently power the motion and on-board sensing systems of the RoboFish limit the operation time to a few hours. Design Process Functional Decomposition The main functions of the RCS was established alongside its subfunctions in the functional decomposition below: Figure 1: Functional Decomposition Chart of RCS Morphological Chart The possible implementations were then considered for the functions illustrated in the functional decomposition chart, below: Figure 2: Morphological Chart of RCS Pugh Chart
Through the use of a Pugh Chart, the possibilities listed above were then put together into a possible concept selection. Criterias were given listed below: Cost Energy Storage Capability Simplicity Meets IP waterproof standard Secures with no risk of damaging fish Shallow Water Charge Time Energy Harvesting Capability Weight Time to complete design Reliability to recharge RoboFish Reliability to attach RoboFish Concept Selection As a result of the pugh chart assessment, the concept that was selected is: Harvest Energy Store Power Charge Robofish Attach Robofish Flotation Method Anchor Automation Solar Panels Lead-Acid Above Water direct connect Fish swims to pole connected to the center of the station, grabs, rises until the charging port on its top connects to one of four ports, guided by one of four funnels Dock Floats Heavy weights Ras Pi Figure 3: Selected concept of corresponding functions for RCS Risk Assessment All the risks of the project were anticipated and prepared for, listed in the finalized risk assessment table below. These risks are sorted by categories as done so throughout the first three phases, but in the final phase, we decided to add a subcategory, whereby the risks are sorted based on whether they relate to Mechanical, Electrical or Software aspects of the group. This way, we are able to resolve said issues with better scope and structure:
Figure 4: Finalized Risk Assessment RoboFish Electrical Systems Though the primary focus of this project is to design and create the RCS, it is important to provide details of the RoboFish as it will be used to connect to the RCS to illustrate that the charging mechanism works. Some changes will also be made to the RoboFish so that this operation is possible. Arduino Mega 2560 R3 The Arduino Mega in the RoboFish is necessary to control the various valves located in its parameters, thus control the fish s locomotion. Both the code and the wiring were designed by the previous team, and no changes were made to it, though minimal changes were made in the earlier stages to test for certain factors (thus said changes were reverted back to its original state). Raspberry Pi 3.0 The Raspberry Pi microcontroller in the RoboFish is required to utilize the Raspberry Pi Camera located on the front of the fish, above its jaws. The pi is programmed to process the image to locate red objects (the pole on the RCS) and communicate with the Arduino Mega to move towards said object. The code did not function at its original state, thus major changes were made by our group. The software flowchart of the functions of the updated code is provided below:
Figure 5: Software Flowchart of Raspberry Pi Image Processing Figure 6: Example screenshot of Image Processing Code
As seen above, the code performs 9 total processes of image processing in order to detect the red hue on the pole. The RoboFish can then center onto the pole and move towards it through its motors. Waterproof The RoboFish was claimed to have been waterproof when passed onto our group, but was not when tested. Therefore, the RoboFish was remade to be waterproof during the first two phases of MSD1. RCS Electrical Systems Grape Solar Photovoltaic Module As discussed in the feasibility analysis of the energy harvesting, two solar panels are required for the RCS. These devices are then connected to the Solar Charge Controller to manage the onboard 12V Lead-Acid Battery. Figure 7: Schematic of Electrical Charging System
RCS Arduino, Switches and Motors The RCS has an arduino pro mini on it, which is to be used to detect electrical connection at a port, secure the connectors against turbulence via microswitch triggered stepper motors which screw the connectors together, and provide emergency disconnect options. The schematic is provided below: Figure 8: RCS Arduino Pro Mini Schematic Fish Side Controller The Fish side controller is always off when charging connector is connected. When powered (i.e. the fish is plugged in to charge), the Arduino listed above is powered, which then controls some transistors or relays (details still to be concluded) to disconnect the load (i.e the rest of the RoboFish electronics) from the battery, and enables the charger. After a set amount of time, the charger is disconnected, the RoboFish electronics reattached and the RoboFish disconnects from the RCS:
Figure 9: Fish Side Controller Schematic RCS and RoboFish software for Charging System Figure 10: Fish-Side Docking Sequence Software Flowchart
Figure 11: RCS-side Docking Sequence Software Flowchart CAD Mode Figure 12: CAD Side View of RCS Design
Figure 13: CAD Isometric View of RCS Design Bill of Material (BOM) The BOM is split into two sections: one contains the materials needed to construct the RoboFish Charging Station and the other contains materials that were used to modify the RoboFish in order to implement our charging system: Figure 14: BOM for RCS
Figure 15: BOM for RoboFish RCS Feasibilities Feasibility analyses have been performed on all major systems of the RCS design. This section describes all analyses performed: Buoyancy Question: How much is the RCS expected to weigh? Answer: 185 lbs Question: How much buoyancy force is needed to keep the RCS afloat and what contained volume of air is necessary to provide that? Answer: 35.9 gallons of enclosed air to float 300 lbs
Energy Harvesting Question: How large of a solar panel area is required to charge the Robofish once per day on an average Rochester day? Answer: On an average day.437 [m^2] is needed
Question: A solar panel was inherited from a previous team. How much power is it capable of harvesting on an average Rochester day? Would two panels harvest sufficient energy to charge the Robofish once per day? Answer: The panel produces 50 [Wh] in a day worst case (rainy, cloudy, in the shade) and 200 [Wh] in a day during an average cloudy day. Two panels will be more than sufficient to harvest enough power to charge the Robofish once (148 [Wh]) during an average day. Docking Question: Can a passive guide mechanism be designed that can guide the Robofish into a docking port if it floats up into it? Answer: Test with miniature scale model improves confidence but is not conclusive due to differences in fish geometry. Figure 16: Scale Model of Guide Design and RoboFish Question: If the guide mechanism is not capable of docking the fish from any docking orientation, can a rotating fixture be used to move the fish into place? Answer: A small scale prototype was made that can likely perform this function.
Attachment Question: Can a prototype connector be made that is capable of accepting the Robofish if its angle is not perfectly aligned? Answer: Yes, the connector can accept a mate that is at least 20 degrees rotated about its height axis. Figure 17: Prototype Connector Design (Isometric View)
Figure 18: Prototype Connector Design (Installation Orientation) Question: Can a prototype connector be made that is capable not shorting when connected after emerging from underwater? Answer: Yes, the connector did not short when tested wet. Structural Stability Question: Can the station remain stable when subjected to small tilts? Answer: Yes, it can remain stable to at least 10 degrees of tilt (likely more but not confirmed). Robofish Feasibilities
The same analyses were performed on the existing RoboFish: Image Processing Question: How far can the RoboFish see underwater? Answer: About 5 meters Buoyancy Question: How much force is needed to completely submerge the RoboFish? Answer: The fish could not lift 5 lbs at full buoyancy Conclusions and Recommendations Since this closure plan only includes details of the project up to the end of MSD 1 (hence the end of the design phase only), we as a group are confident on the design of the charging mechanisms on both the RCS and the RoboFish. We are, however, still concerned with the current state of the RoboFish. It was passed onto us with incomplete wiring and wiring diagrams, making it difficult to test the RoboFish. One of the first tasks for the next stage is to re-wire the RoboFish so that it is both neat and ready to be improved upon for future assignments. At the moment, all wiring in the main box of the RoboFish has been disconnected, and is in the process of rewiring. It is highly recommended if possible to provide a larger box for the electronics, as it is expected that once the new electrical systems for the charging capabilities have been added to the box (such as the Arduino pro, switches and relays), it will clog the box up enough to make it difficult to make any further changes without having to disconnect. The limited space also makes it difficult to neaten said wiring. While working as a group, it is also recommended to split the group into different aspects of the project, that is: Mechanical, Electrical, and Software systems. Doing so will help maximize engineer s strengths within these fields. Acknowledgements Our sincere thanks goes to Dr. Kathleen Lamkin-Kennard and the Boeing Company for providing us the opportunity to work on an on-going project with real world applications and development. Dr. Lamkin-Kennard has been communicating with our team tirelessly to ensure customer and engineering requirements have been completely satisfied. We would also like to thank our project guide Mike Blachowicz for sharing with us his expertise in engineering designs and his wisdom in the importance of process and iterations. Finally, we would like to thank Kate Gleason's College of Engineering for providing not only facilities needed for the project design and implementation, but also professors who have passed on invaluable knowledge in embedded systems, classical control, and fluid dynamics among others, without which, we would not have been able to facilitate the design problem.