Team Vision for System-Level Design PhaseOur team aimed to create a system-level design of the ComQuaT system during this phase. This process included functional decomposition, morphological analysis, Pugh analysis, and preliminary testing plans as well as updating our benchmarking, risk assessment, and long-term project plan. Our team achieved these objectives and were able to make several preliminary design decisions.
Functional DecompositionThe ComQuaT system will be completing an expedition of ten feet or more in a straight line across even ground. The robot will be deployed from an RC vehicle and connected via a single umbilical cord. The expedition will be a walking motion powered by the fluid articulation chambers of the robot's limbs. here.
BenchmarkingExisting systems were found and compared to each other by critical specifications, providing insight to our design process. here.
Codename: UC-MSDThis system is based firmly in the quadrupedal robot built at University of California – San Diego in May of this 2017. It includes a chassis that angles each leg downward about 45° with 90° between each leg. The legs are composed of three chambers with eight units that form a sinusoidal pattern on each one. The tether protrudes from the top of the robot and attaches to the RC platform in the front and secured with magnets. The leg chambers are filled with ambient air, compressed by micro-compressors mounted on the RC platform.
Codename: JellyfishQuadrupedal robot with leg chambers that are designed to inflate quickly and articulate significantly with low volumes and pressures (see P17027). Robot would be secured to crane arm with magnets and removed from or placed on the RC for deployment or reattachment. Air supply would be compressed into a paintball tank prior to expedition(s).
Quadrupedal robot with leg chambers that resemble the UC-SD robot, but with more angular ring sections to collapse deflated leg material with more completeness and repeatability. Robot would be grasped by an articulating crane arm and removed from or placed on the RC for deployment or reattachment. Air supply would be provided by micro-compressors.
Codename: HulkBotQuadrupedal robot with leg chambers designed based on the UC-SD robot, with fluid powered by a large compressor mounted on the RC platform. The compressor is powered by large batteries mounted on the RC as well, making the RC much larger and heavier than other concepts. The robot is magnetically secured to an actuating crane arm to remove it from or place it on the RC for deployment or reattachment.
Inflatable GripperDuring concept selection it became clear to the team that with any of the above designs we would leverage the design of the robot’s chassis and the design of the tether and deployment systems to make the robot also function as a gripper when onboard. When not deployed the robot will be mounted on the front of the RC platform in its deflated state. The robot’s legs may be inflated without deploying the robot to articulate and grip objects.
Feasibility: Prototyping, Analysis, Simulation
From the functional decomposition and preliminary design concepts, the necessity for a microcontroller has been identified to control the electrical components of the system. The following microcontrollers were compared to make a decide which controller would best suite the needs of the project.here.
Based on this information, the best option for the application seems to be the Arduino Mega. The Arduino Mega has the most input and output pins which will be beneficial while controlling multiple valves for the articulation of the legs. The Arduino boards are also programmed using C as a language. This is a programming language that the group is collectively most experienced using. The Mega also has a reasonable price point. For these reasons we expect to use the Arduino Mega in the design.
The necessity for a pneumatic supply and power supply were also evident in preliminary designs. The following table compares several options for air supplies and power supplies. here.
The concepts shown in the Pugh chart were then analyzed for their feasibility by examining the power, pneumatic, weight, and cost realizations. The following table shows the calculated quantitative requirements for each concept.here.
Because our designs are grounded in the design from UC-SD we are able to utilize the data and analysis published in their paper "3D printed soft actuators for a legged robot capable of navigating unstructured terrain". Below are figures of the analysis for Extension (fig. 8), Bending (fig. 10), and Force Output Comparisons (fig. 11).
Morphological ChartA morphological chart was created to assess the solution space for critical functions, see below. here.
Concept SelectionA snapshot of our Pugh analysis, including the selection criteria used:
- The "UC-MSD"system had the most positive marks
- as well as the fewest negative marks.
- The "Jellyfish" and "Hulk" systems tied for the most
- "Jellyfish" also had the fewest positive marks.
Systems ArchitectureSystem Architecture broken down into six subsystems (Pneumatics, Robot, Electrical, Drive, Tether, and Attachment), and further into a preliminary components list. Interactions between the subsystem components are identified on the right side of the breakdown.
Designs and FlowchartsDiagramming the fluidics for the ComQuaT system is a key step in the development of our product. We iterated designs with team input to make our system as concise, elegant, and feasible as possible. As we select components for the final design of the ComQuaT system this simulation could be updated to model the behavior of the pressure in the leg chambers.
Risk AssessmentThe Risk Management Chart has been updated to now include risks that are relevant to the updates made to the design of the robot. These include a risk involving if the robot acts as a gripper, a risk involving the regulators and valves and a risk involving the tether that connects the robot to the RC Vehicle. A risk involving the robot being used as search and rescue has been removed, as we will not get this far with our design.
BudgetOur budget goes over our allotted amount given ($500). Since this is the case we are looking into a grant from IEEE, this will allow us to have enough money for all of the materials we have found. Another way around this will be using other means rather than the pressure transducers in our robot.
Plans for TestingWe will be testing the fabrication process for our leg systems. Based on an instructable created by user mikey77 called "Soft Robots: 3D Printing Artificial Muscles" we will be 3D printing our leg chamber alpha prototypes (see below) in NinjaFlex material with the help of the Brinkman Lab and coating the prototypes in a a solution of Loctite glue and MEK Solvent.
Design Review MaterialsThe handout that will be available during our presentation can be downloaded here.
Plans for next phaseBy our next review, our team aims to complete a detailed design of the ComQuaT system. This will include the following for each subsystem:
- Requirements Flow-Down
- Complete Schematics
- Demonstrated Proof-of-Concept
- Plans for Testing
|Project Manager||Conor McKaig|
|Lead Engineer||Zach DiLego|
|Primary Electrical Engineer||Cameron Taylor|
|Primary Software Engineer||Sean Bayley|
|Primary Hardware Engineer||Zach Hayes|
|Purchasing & Materials||Marie McCartan|
|Communication & Customer Contact||Jamie Mortensen|