Weekly Updates

In the first week the project was chosen and the team was formed. Bionic Technology Organization is senior design team comprised of Daniel, Nate, Nick, & Phil. Their senior design project is to design and build a biomechatronic prosthetic for above the knee amputees.

Week two, the team continued their project by defining a problem statement, system requirements, and goals. By defining a problem statement and system requirements this allows the team to be bound to certain expectations for the project. The system requirements document can be found in the System Requirements tab of this website.

The team conducted a patent search and literature review to further their understand on mechanical prosthetics, biomechatronic prosthetic, and the anatomy of the human body. This research allows the team to develop a stock of sources to reference when developing their project. The team gathered their findings in a technical document and presented the information to their peers. The document and presentation can be found in their respective tabs of this website.

Team developed individual conceptual designs for the biomechatronic prosthetic. The designs were intended to be high level and prove that the concept is able to meet the system requirements. Each member of the team modeled their own conceptual design in a computer aided design program, created an electrical system block diagram, and a software flow chart to present to the section.

Team developed a group conceptual design for the biomechatronic prosthetic to meet all of the system requirements. After presenting the idea and receiving feedback the team made revisions to the design and began brainstorming system specifications. A specification document was made to outline specific values that need to be met in the testing phase to validate the project requirements. Additional the team revised the style guide to document switch from Microsoft Word to a LaTex text editor.

The team presented the project specification report and received feedback. Created a virtual workspace to organize and assign project deliverables. Continue to make improvements to the style guide documentation and update the specifications.

The Team began a preliminary design CAD model with motors, actuators, and mechanical components they would like to use in their final design. Furthermore, the Team worked on deriving the forward and inverse kinematics for the two degree of freedom ankle. With the project progressing the team realizes they are going to need sponsors and/or parents which began the search for support.

Team BTO finishes version 3 of conceptual CAD. A determination was made that the motors used in design are outside of budgetary constraints so an alternate motor needs to be obtained. The team is considering using DC motors with a 100:1 gearbox rather than the prepackaged linear actuators. The forward and inverse kinematics are almost complete.

The team completed another round of CAD with DC motors. The motors were later determined to be underpowered and the design needs to be modified. Electrical compatibility was considered between the motors, battery, and gearbox. Forward and inverse kinematics have been completed. The style guide has been updated.

CAD design for the knee has been completed using a new Maxon motor that has 200W of power. Compatibility between the motor, gearboxes, encoders, drivers, and the control system is being researched and determined. The latest ankle CAD is almost complete pending updating the motor being used to drive it.

Team completed CAD for the entire leg with a three step 100:1 timing-belt pulley system in the knee and a 100:1 gearbox for the ankle. Due to budgetary constraints the team is considering using motors with less power and changing the project requirements to meet the specs of a less powerful system. A simulation has been completed in MATLAB using dummy values for materials and weights that have not been determined yet. Electrical schematic work is underway pending completion next week.

The team is making edits to CAD and working with machinists to finalize the gear drive design. Rapid prototyping of the foot has led to additional design updates. Electrical considerations are being calculated and completed CAD drawings of the electrical components are being added to the final design.

This week the team is finalizing on/off ground design simulations and working to finish the final draft of the preliminary design document. The presentation is outlined as a rough draft and research is being compiled to reference in the final documentation.

The team is editing the final preliminary design document and working to finish the presentation for the panel of faculty. Presentation rehearsals are being conducted and requirements are being reviewed to meet all of the project deliverables.

The team is making edits to the CAD design in preparation to 3D print parts for dry fitting and initial stress testing. The team met virtually and updated the project Gantt chart to reflect the changes to the project and set milestones for the project throughout the semester. A rough draft outline for the test plan was created in Overleaf.

The team cleaned up the team drive and continued to work on the final drawing package and stress analysis. The test plan was started and continued work was put into programming the electrical system. A parts order was placed so that the team can begin building as soon as in person instruction resumes in February. Website has been updated with the team bio.

The team placed additional parts order and submitted tickets for 3D printed parts. Testing of the electronic system has commenced and multiple subsystems are communicating as intended. Machining of the motor axles and joints is underway. The foot sled spring rod assembly has been machined and the retaining washers have been water jet cut.

Machining work is continuing to take place and the team is moving towards a prototype assembly of the leg. The team acquired parts with different dimensions from the original plan so another round of CAD revision took place to account for the new sizes. A test ankle adapter was printed that will eliminate the need for a motor mount since the motor dimensions were incorrect from the manufacturer. Additional 3D printing will take place over the weekend. The team is projected to have a prototype dry fit by next Friday.

After four rounds of 3D printed parts a number of updates had to be made to the design. Additionally, tolerances of the steel key stock and the 3D printed parts have led to the redesign of 10 components of the leg including the ankle sled, helical gears, knee/ankle motor mounts, and the spring connection adapter. The spring is also not long enough to achieve the designed compression at the intended angles so a preload adjuster tube was designed to adjust the spring preload. Currently we are one week behind schedule due to the multi-day lead times for all updated 3D printed parts.

With the aluminum shaft machined and adequate amount of 3D printed prototype parts printed, the we were able to create a prototype of the leg assembly and test the motors. The following video shows how this prototype performed:

After building version one of the leg the team determined that the horizontal helical gear had excessive play and a redesign was required. A cross notched section was extruded from the back of the gear and a matching slot to fit in the knee and ankle motor mount. Once printed it was determine this created a more rigid connection and eliminated the horizontal backlash from the shaft. Additional fillets were added to the motor attachments and dimensions were adjusted to achieve tighter tolerances around the meshing gears. The motor shaft adapters were retapped to receive larger set screws which needed to be ground down to fit in the motor mount. This allowed for a tighter fit of the motor shaft adapter onto the gearbox output shaft.

After multiple tests resulting in stripped gear teeth it was determined that resin printed gears would be a better option. This theory was put to test and was partially successful. The resin gears were able to survive the forces of operation but after multiple cycles the gearbox motor adapter shaft became loose de-meshing the gears and causing pitting in the gear teeth. A redesign was drawn up to mount a bearing directly underneath the motor side gear to transfer the bending moment of the gear box shaft to a bearing imbedded in the motor mount. The team is currently waiting on 10+ parts to finish printing at the rapid prototyping lab and an order of parts to be placed before further testing can occur. Progress was made in the electrical and simulation side of the project with a breakthrough in the motion capture code. The team was able to detect three separate frames on a human leg and plot in MATLAB the rotations relative to the body fixed frame on the hip. This data will be critical to teaching the machine learning algorithm in tandem with the EMG sensors.

Work was done to assemble the leg in its entirety. Reduction of the backlash in the gears was successful, however repeated cycles of movement worsen its effects. Further work will be needed so that the electronics can be mounted and the wiring done.

Soldering was done on the EMG sensors so that they can be used more easily.

This week, the final fabric layers were added to the composite foot and it was cut out out to shape. Further machining will be needed to get it ready to attach to the leg. Additionally, the electronics were mounted to the leg and much of the wiring was completed.

More work was done on the motion capture, but further testing will be needed to get the data desired.

This week, progress was mainly made to improve the workflow of the project, while there were steps taken to bring this project closer to completion.

More mocap testing was done, but it was found that the wires produced too much noise which got in the way of collecting useful data to map muscle movement to joint configuration.

A test stand was made using a mic stand to suspend the leg, which also proved to be useful as a platform to work on the leg. In addition, the carbon fiber foot that was finished in Week 9 was attached to the leg.

For the code, the leg can now be controlled through MATLAB so that control logic can be implemented and tested.

Lastly, the positions of the PCB, motor controller, and Jetson Nano were rearranged to be easier to manage when it came to wire management.

Much progress was made to get the leg ready for testing. New parts were 3D printed to fit the design better, bump switches were implemented in the design and made functional in the code, and steps were made to limit backlash in the gears as much as possible.

For the code, it was found that Python was a better choice to run the Jetson Nano reliably, as it supported by Nvidia for the board. As such, some time was dedicated to make the bulk of the code run in Python. With all these advancements, the leg can now be calibrated thanks to the bump switches, and controlled with PID controllers in the MATLAB code.

Lastly, wires were cleaned up and the leg was made to look presentable with the addition of the electronics cover.

The leg is now in a state where much of the specifications can be put to the test, which the teams plans to do throughout Week 12.

This week, all efforts were set on finishing the final motor power test and on continuing progress on the test document itself.

All tests were completed and the test document is at a near complete state. The last part of the test document is set to be completed by the end of Week 13 so that the team can focus on preparing the final report and presentation.

All focus was put into finishing the test document. Once the document was finished, the next task was to start the final report and presentation. It was decided that part of the presentation will include a live demonstration of the leg being controlled through mocap data, so plans were made to set this in motion.

Now that the test document is completed the team is working on finishing the final detail design report and presentation. Photos have been taken of the assembled leg and the CAD assembly has been re-mated after numerous revisions throughout the course of the project.