Customer Handoff & Final Project Documentation
Table of Contents
Team Vision for Final Demo and HandoffDuring this phase, we proceeded to troubleshoot problems that arose with the system. The two most notable dilemmas faced in this phase were problems with the microcontroller and the servos. The microcontroller was having difficulty interfacing with all the electronics, so we changed the type of microcontroller, along with the code in order for proper implementation. The servos also had problems working with the power board, so we researched and purchased new servos for the system. The mechanical team finished the all the designs and machined or 3D printed all of the housing for the devices.
Final Image Of System
Test Results Summary
The team has run in to many problems during this phase which has resulted in a delay in testing. For example the pulley system that was originally designed was not strong enough and the belts would slip. This required us to redesign our system. On the electrical side, there were a few hiccups which prevented us from finishing all of the testing. This was because some of the system peripherals(Lidar, accelerometer, display, ect) stopped working. We were then unable to complete some of the tests because we had to get the system working again first.
A link to the live document containing the completed test plans here
Risk and Problem Tracking
The newest and major problem we are having is the changing of microcontroller from an Arduino to a teensy. This change involved making a new surface mount power board and housing to mount the device to the bike. This implementation was a success and the electronics connect together fairly well. The second problem was that the servos kept getting fried, so they had to be replaced with a new, better rated, type of servo. This implementation was also a success.
A link to the live document containing the Problem Tracking Chart here
Final Project Documentation
- A link to the team's technical paper can be found here.
- A link to the team's final Poster can be found here.
CAD/SchematicsAll of the CAD files that were created in the design process as well as the schematics that were used can be found by going to the Detailed Design Documents folder on the teams SVN folder. This location houses both the mechanical and electrical models and schematics.
Final BOMThe updated Bill of Materials is shown below.
A link to the live document can be found here
Assembly InstructionsA .PDF file of the Installation manual can be found here. An excerpt is shown below:
A .PDF file of the operator's manual can be found here. An excerpt is shown below:
The Arduino code used to program the Teensy 3.6 microcontroller used in the current revision of the hardware can be found here.
Please note, this is a working revision of the code as of 12/13/2017. The teensy software library will need to be installed in order to flash the board through the arduino IDE.
Recommendations for future work
MechanicalFirst and most important is that the system needs to be waterproofed much farther beyond the current state. The only parts on the current design that are water resistant are the control switch and the the LIDAR housing. This is a key requirement of the system that needs to be met if it is to be fully tested and implemented in the future.
The mechanism for adjusting the suspension does not work as well as it could. The current mechanism utilizes a belt and pulley system that requires large tensions on the pulleys to keep the belt from slipping when the servos turn. This tension applies a harsh moment onto the shafts that the pulleys are mounted on. If this mechanism were to continue to be used the shafts may fail prematurely due to the fatigue that comes from the moment placed on the shaft via the belt tension. If a locomotion mechanism similar to how a train wheel turns were implemented there would not be a constant moment on the shaft. This would potentially increase the life of the shaft. The locomotion mechanism would also work in more environments than the belt system since it would be a rigid bar and not a pulley that relies on the friction between two surfaces that can potentially change if the environment changes. (this is illustrated in the technical paper)
Some of the handle bar mounts like the LIDAR mount could be made of aluminum instead of 3-D printed plastic. This would help to cut down on the vibration that is introduced into the system because of the flex in the plastic parts. An aluminum enclosure and mount system would not only increase the crash resistance but also reduce the noise in the LIDAR. This would in turn make the proactive sensor more accurate and tighter threshold could be used.
Electrical• Create a PCB for the power control board to allow for a smaller board to be used. By decreasing size, less space is consumed on the frame of the bike.
• Add power saving to the code to shutoff power to devices when in manual mode. This will allow for a smaller battery which would decrease weight while maintaining the same operational performance as the 7.4V LiPo battery being used currently.
• Put shielding on the wires to prevent the wires from breaking from its connections
• Use separate I2C lines to each of the peripheral devices. Since the teensy board has 4 I2C lines, the exposed board that connects the I2C lines can be removed, reducing the weight of the board and eliminating the need for housing the board.
• Redesign the power board to work with a teensy microcontroller
The last-minute change to the Teensy 3.6 controller based on the NXP K66F controller allowed for a substantial addition of hardware functionality. power saving functionality should be added in future revisions of the hardware board that includes the power MOSFET and switching relays for the peripheral devices. Adding a few buttons to the outside of the main enclosure would also benefit the project, as this could select between power modes.
Presently, the LIDAR and accelerometer devices are polled within the main loop, however as the Teensy has multiple I2C peripherals, it would make sense to use a hardware timer and interrupt based I/O scheme to sample the sensors independently at their maximum sample rate. This would allow for a more streamlined approach to the autonomous mode, and faster reaction times from each of the sensors.
Originally a customer requirement that the team believed we wouldn't be able to commit to, datalogging the sensor data into a file on the SD card is now possible. The graph below shows the accelerometer and LIDAR data for a simple test, dropping the bike from a height of 3 feet with no rider.
Plans for Wrap-up
Devin Cooley Plans For Wrap-Up1) Support the electrical and Software team in getting the bike to a testable state.
2) Help with testing the bike.
3) Hand the bike off to the customer and use feedback to make minor improvements before the end of the semester.
4) With the help of Lorenzo Write the Assembly Manual so that future teams. know how to properly install the suspension system.
5) Have some beers with the rest of the group.
Andrew Lints plans for Wrap-up1) finish implementing the datalogging features on the new microcontroller, and start to look at the collected data to see how we could use filtering methods to get better responses from the servos.
2) start to clean up the bench area and make sure that all parts that are necessary to continue work on the bike are collected.
3) Maybe look into different methods of creating a top for the switch, and have the construct mill a new pcb for the new switch (as the right toggle on this one has stopped working reliably).
4) Look into slimming down the wiring and housing on the bike a bit to help with appearances and waterproofing.
5) Go for beers with the best MSD II team this semester.
Lorenzo Nunez Plans For Wrap-Up1)Complete pressing finishing tasks such as covers and 3d prints.
2)Help teammates with wrap up documentation.
3)Put designs in correct folder.
4)Help Devin with the instruction manual.
5)Help Andrew with those Beers!
Kristina Shillieto Plans For Wrap-Up1)Help to finish all the documentation.
2)Finish updating all the Edge pages.
3)Finish the BOM and Problem Tracking Documents.
4)Clean up Workbench.
5)Enjoy my last week of college by getting beers with an amazing MSD group.
Nemanja Drobnjak Plans For Wrap-Up1) Assist in the completion of the write up of documentation.
2) Ensure all relevant files are inputted on edge.
3) Flip the final image of system picture.
4) Ensure cleanliness within the P17214 EE senior design room Workbench.
5) Drink with superb MSD group.
Zachary Law Plans For Wrap-Up1) Help clean the workbench.
2) Ensure all relevant files are inputted on edge.
3) Decide who gets to keep the final product.
4) Make sure locker is clean and surplus supplies are given to MSD.
5) Have a few drinks with the team to celebrate the end of MSD and college.