We survived!
After much relaxation, I (Edmond) can finally say...
After much relaxation, I (Edmond) can finally say...
The past two days were very eventful. Early morning on December first, Jake announced the HSM should allow the robot to find ammo and the enemy robot. Unfortunately our mechanism to keep the motors in place failed when one of the motors inexplicitly dropped and the wires became too damaged the motor doesn't work. Unlike every other team's motors, ours have the wires exposed from the side not the front, and securing them with our motor holds damaged the wire sheath. Now we pay the price of that. Jake tried to fix the motor by soldering the wires back but they only go half as fast and he made them in a worse state by opening it. We spent the first half of the day borrowing motors from Max and integrating them onto our assembly. While the motors do work, they seem to be slower than the motors we had before, despite being rated for 12V at 81 rpm. After that Luis continued to work on the FSM, but now he has to revise much of the code to account for the slower motors.
We spent much of the day getting the hardware fixed and then starting this afternoon worked to reach closer to checkoff. Luis spent the early part of the day getting the Finding enemy FSMs working, but for unknown reasons the motors were only run for a couple minutes after plugging in a fully charged battery. We spent the whole morning figuring out why, but by noon Edmond was the only one around. He and Luis worked with Max the TA to determine the cause of the problem. We initally thought it could have to do with our daisy chained regulators or a current spike somewhere in our electrical system. We found out on the internets that daisy chained regulators are common, and even Max encouraged it. We let the Uno and our tape and bump sensor perfboard run from the battery for many minutes, and then when running the motors, the "Going to sleep" message came up in seconds. Replacing the battery does nothing to remedy this.
Spend another long day in the lab.Spent a lot of the night in 138 lab, on the field testing the bot. My hope was to get the findingEnemy sub state machine completely working. Unfortunately it took me a lot longer then I hoped it would.
Spent a good amount of the the day in lab. Made some good progress. With Line following working. Today we worked on getting the bot to get on the the black tape "T", and get some ammo. This involved adding two more sub state machines, Our third levels, FindingTSubSubHSM and AligningTSubSubHSM. Im working on documenting the state machines in Umlet and will post once they are done. Essentially these state machines wall ride the depots, until it gets to the "T" and then AligingTSubSubHSM works to get lined up perfectly to release the ammo.
Happy thanksgiving! We begin the day at midnight when Jake created a milestone for the FSM. After a couple hours of testing, he created a FSM that allows the robot to successfully find the tape by going straight at it, aligning itself with the tape, and then following the tape by rapidly swerveling back in forth on the line but still going straight. Later today Luis and Edmond worked on the next parts of getting ammo, finding and aligning to the T. We both decided that they should be in their own sub level FSMs to simplify the HSM.
Today we have transitioned from assembly to programming our machine and also started to design the ball launcher motor! We should be on track to finish by checkoff next week.
Over the past two days we finally took the steps necessary to integrate most of our robot except for our launcher funnel and beacon holder to start focusing on just the HSM. Yesterday, we mounted our tape sensors and determined the new hysteresis values. Most of the tape sensors had similar hysteresis values. At first we had a lot of trouble figuring out why two tape sensors would have the same behaviour, we can place one of them over white and the other over black and both would be outputting its high voltage or both would output its low voltage. We resolved this problem by ensuring each sensor was plugged in to its own IO port and not to ground, our ribbon cables are not labelled which lead to this confusion.
From Edmond:
Today we just walked on tape sensor placement and prototyping. First we edited our solidworks model of the motor holders in order to make sure they don't interfere with the rod support holes. We also editted our sensor base to shortened the distance between the two motor holders on the left and right halves of the base. That meant the side panels that supported them are no longer needed and now space is freed up for tape sensor placement. We then created a new assembly to make sure the motor holders and the sensor base mated corrected. We laser printed foamcore and MDF models of the assembly parts and found this setup will have to work because it was a tight fit for the motors and the motor holders and hopefully we didn't mess up the wires.
Progress in the past few days have been going on slowly because we are busy with other assignments due before thanksgiving but we will do whatever we can and work during the holiday.
My first blog post. The last few days were spend locking down the mechanical design of our robot. By thursday we believe we came up a with solid design, and are ready to move forward with it. The hope is by Friday night can get everything mounted and get all our sensors working. And Continue working at adding capabilities to our HSM, as things are added to the bot.
We did several things today as we looked into tape sensor placement. Yesterday Luis brought up ideas to fit the sensor base for the tape sensors by modifying the box that holds our motors.
Today and yesterday we worked on getting our bump and tape sensors to work and started to integrate them with the rest of the robot. Here's a list of what we done:
Today we made significant progress by printing our base in MDF, attaching our motors, and finally getting it to move with our test harness code!
Following our first laser cuts we decided to focus just modeling a base that we can quickly print and assemble. Luis and Edmond decided that the motor mounts should be ditched in favor of motor boxes. Neither of these "mounts" serve the purpose of securing the motors, they are meant to support a base level for the tape sensors since they need to be an inch off the ground. We also decided to seperate the wheel mountings and the board mountings, that is they belong in two seperate bases but are seperated by a short tab slot divider. This allows room to mount our Uno32 stack and daughterboards. We also modified the screw hole for the bumper mechanism to support the screws we have.
Last night we finally got a chance to laser cut our base in MDF and Foamcore. Here are our lessons we learned:
Today we had our weekly meeting with our mentor to determine what we have done this weekend and we need to get done by next week. We do have our circuits built and we are currently writing our test harnesses to communicate with our sensors. We are also working to get our sensors on perfboards, so far we have our tape sensors and one beacon detector on perfboards. We have been able to run our motors with our H-Bridge, and we have most of our parts that we need.
We have been moving along slowly but have made steady progress. We originally wanted to start laser cutting our foamcore robot model by today but we are currently fixing our Solidworks model to ensure all the individual parts mesh together in a way that we can construct the robot when we laser cut our foamcore board. There are quite a few hole mounts we need to create to account for components that we want to secure onto the base and sensors we want to mount towards the floor. We also have to make sure our robot is less than 11 inches tall (not including the beacon), currently it is a couple inches above that limit. The bottom base have holes for the pipe mounts, bumper supports, caster wheels, and 2nd floor attachment. We designed our ammo plunger with a sharp end. Since the ammo depot requires very little force to be pushed to let the balls out, we decided the best way to do so is to apply as much pressure at one point. Until we print out ammo plunger we can't mount our bump sensors just yet.
Among the parts we salvaged from a Xerox Laser Printer are a few solenoids. When brainstorming ammo collection we decided that solenoids were a better alternative towards keeping ping-pong balls in place instead of a servo because of its push and pull motion of a leveler. In this video we demo operating the solenoid by sinking and sourcing current.
Today we successfully passed the design check-off of our project! We had to complete the following tasks:
It has been a couple days since we meet our mentor who gave us some suggestions on how to handle the project. He lend us motors and gave us stretch goals to work on by the end of this week. We made a [Gantt chart](http://iamtechknow.github.io/118website/#gantt) to visualize our goals for next week. We plan to have an autonomous drivable robot by next Saturday!
Our robot shall be known as Admiral Uno32Stackbar!