Weekly Log.
Week 1: Commence Operations
This was our introductory week. We decided to hit the ground running by prototyping a hovercraft base and beginning to lay out the groundwork for our communication. Dongao and Bobby prototyped the first of our many hovercraft bases, and Kuangyan begin very initial coding. Phillip was busy on the communication team at this point, reading up on Zigbee communication protocol. By the end of this week a rough draft of the communication protocol was completed as were several potential hovercraft bases, using different skirt material such as plastic sheet from Room 36 and a child's inner tube. Below is a video of our initial prototype. We also prototyped a balloon popping mechanism that utilized a hot wire. It was super effective.
Then things really got underway. Bobby and Phillip visited an amazing hobby shop called J&M Hobby House in San Carlos. If you ever need to make an RC anything, go there. There they got expert advice, as well as propellers and motors. They also bought several yards of ripstop nylon that would be used for a more finalized hovercraft skirt. Many more rough hovercraft prototypes were made, utilizing foamcore, duron, and different variations on a soft skirt.
This was our introductory week. We decided to hit the ground running by prototyping a hovercraft base and beginning to lay out the groundwork for our communication. Dongao and Bobby prototyped the first of our many hovercraft bases, and Kuangyan begin very initial coding. Phillip was busy on the communication team at this point, reading up on Zigbee communication protocol. By the end of this week a rough draft of the communication protocol was completed as were several potential hovercraft bases, using different skirt material such as plastic sheet from Room 36 and a child's inner tube. Below is a video of our initial prototype. We also prototyped a balloon popping mechanism that utilized a hot wire. It was super effective.
Then things really got underway. Bobby and Phillip visited an amazing hobby shop called J&M Hobby House in San Carlos. If you ever need to make an RC anything, go there. There they got expert advice, as well as propellers and motors. They also bought several yards of ripstop nylon that would be used for a more finalized hovercraft skirt. Many more rough hovercraft prototypes were made, utilizing foamcore, duron, and different variations on a soft skirt.
Week 2: Like Fire
The beginning of Week 2 saw a Spectre mechanical base completed in CAD. A skirt was created as a separate assembly out of nylon and foamcore. It worked well, and would ultimately be the final assembly of the mechanical base.
Communication seriously got underway in Week 2. Bobby and Phillip worked on Spectre (hovercraft) code that would receive a packet of data from the MI6 (remote control) and analyze it to control the motors and lift fan. Dongao and Kuangyan wrote the MI6 code that would read in a potentiometer value and send it to the Spectre.
The end of week 2 saw our our first checkpoint hit on May 12th. For the first checkpoint, Team 11 had decided they would have a fully constructed mechanical base and be able to send commands from an MI6 to make the hoverraft drive. The communication did not yet have to conform to the meticulous protocol developed by the Communication Team that would allow each team's MI6 to control every other team's Spectre. The mechanical base had largely been constructed in the past week, with the motors and propellers from J&M Hobby House (seriously, check them out) mounted to a duron frame. Communication had been worked out towards the end of week 2, with packets from the MI6 being received by the Spectre. Despite a few bugs, the code on either side came together pretty quickly.
The checkpoint was met. We were able to drive ten feet, turn around, and drive back ten feet. It was a great success, though the challenge was very far from over.
Week 3: Second Checkpoint
The second checkpoint was six days from the first, on May 18th. For our second checkpoint, we had to drive forwards and backwards and this time, conform to protocol. This means we had to both demonstrate that our MI6 could control other Spectres and that our Spectre could receive and understand commands from another team's MI6. Initially we had several bugs with pairing/unpairing. Once we fixed bugs in Spectre code causing us to automatically unpair and fixed code in the MI6 that was also leading to unpairing shortly after pairing, we were off. The checkpoint was met on schedule.
Three days later was the Project Preview, effectively the third checkpoint. For this, we had to have both our balloon monitor in place and the code written, and a balloon popper mounted. Dongao and Kuangyan were busy working on a phone app for the MI6 and a bluetooth connection to allow it to talk to the MI6. Bobby wrote the balloon monitor code for the PIC, and Phillip wrote the balloon monitor code for the Tiva. It worked, and then it didn't work - we were getting our packets all out of order. We eventually realized our many print statements were getting in the way. Once we removed them, comm started working perfectly. Remember, print statements are a great tool for debugging, but they are very dangerous in communication code where timing and interrupts are being used heavily. Be wary.
Week 4: The Final Stretch
Nearing the end of both this project and the tumultuous, rewarding, at times devastating and exhilarating journey through ME218, the final stretch was far from the end. Bobby and Phillip set about making a more robust balloon popper out of duron. They also spent an evening pretending they were in kindergarten eating ice cream bars and making a mess with paper mache. Meanwhile, Dongao and Kuangyan finished up the app and made the MI6 look awesome. Checkoff came and went surprisingly painlessly, then came game day the next.
Phillip took the morning off to go buy pokemon shirts in Fremont. It was worth it. The rest is history.
So long ME218. Thank you for everything.