What is F.I.R.S.T. Robotics?
Every January FIRST releases the details for that year’s unique competition. Starting that day each team has 6 weeks to design and build their bot. At the end of build season each team’s bot is sent in for inspection. Not long after, competition season starts. Local competitions select the teams the qualify for Regionals. Only a handful of teams from each region qualify for World.
In addition to scoring frisbees, climbing a 14 foot pyramid was the challenge for this year.
Among the many things that I had a hand in designing, I took lead on a few.
What I did
I designed the shooter entirely. I started by using the highest power motor allowed (AndyMark 337W CIM). I calculated optimal gearing ratios that could increase the exit velocity of the Frisbee without stalling the motor. A high normal force was required to keep the wheel from slipping on the disk. In order to achieve that, I graphed the force needed for varying levels of compression. I found 13% compression was optimal accounting for energy losses in plastic deformation. After CADing the model, I built a prototype. Of all of the teams our shooter was one of the most powerful, accurate, and reliable.
Our unique method of scaling the pyramid involved driving up the corner. We used claws that firmly clamped the wheels to the pole.
Calculating the immense forces needed to keep the drive wheels from slipping, led us to use high powered pneumatics to actuate the claws.
Watch closely as claw retracts when encountering the intersections.
Two teams, tree bots each have 2 minutes to collect scattered basketballs and escort them into your team’s hoop. Traversing a 4 inch high beam at half-court opens up offensive play, simultaneously stealing their balls. In the last thirty seconds bots may attempt balancing on the ‘bridge’ with their teammates for extra points.
Due to motor limits, it is often better to specialize your bot with a specific skillset. Ignoring some of the obstacles allow for a less complicated bot. However, clever design allowed this bot to do it all.
The half-court beam brought on the most challenges.
Using one motor to power three accessory wheels required unique gearing, transferring torque across the center of the bot.
What I did
- Used high power motor (BaneBots 550)
- Stored rotational potential with 2 high grip wheels
- Used tachometer to measure when wheels were at max speed
- Tested various compressions for maximum exit velocity
- Designed and constructed carbon fiber housing
- Designed in Solidworks
- Used ultrasonic proximity sensor to detect when a ball had been picked up
- Calculated lift distance in order to store 3 balls
- Integrated the lift gear into shooter housing to deliver ball to shooter correctly.
- Mounting motor optimally to lower center of mass.
- Designed belt tensioner in Solidworks.
Each hoop’s backboard was lined with retroreflective tape, allowing expedite the aiming process, autonomously lining up the perfect shot every time.
Lowering the center of mass was critical. Heavy components were designed to be mounted as low as possible. Composite materials kept mass minimal wherever possible.
Points were accrued by passing the ball between allies. Scoring the ball then awarded the accumulated points.
This year required teamwork. Unable to control your ally’s ball handling, we expected to be chasing around lose balls that had been dropped. My focus was on ease of ball retrieval. The final build would collect any balls that even got close.
What I did
I took lead on the ball manipulation. After calculating the forces needed to accelerate the ball to a reasonable speed I ruled out the use of wheels. Instead, I explored storing energy using elastic. I mounted the assembly at a 45% angle for maximum range. I implemented a pneumatic system to enable two shooting angles for when the bot was closer to the goal.
The ball pick-up mechanism was also my project. A lot of thought went into this but I’ll give you a summary. The assembly needed to extend and retract quickly, so, I chose pneumatics to power that motion. Using some geometry, I calculated the force needed to retract the arm from the extended position. Also, the “beater bar” used two different wheel radii to pull the ball in. The 2 different sizes were used because they would have varying angular velocities, this was used to steer the ball into the bot.
Designing an affordable, omnidirectional drive train was our biggest breakthrough for the team.
The team worked together to iterate on the last years design to make the assembly more cost effective and efficient.
The bot’s new agility gave us the maneuverability to avoid frequent standoffs on the playing field.
Each wheel being independently controlled paired with clever code, this drive train allowed the bot to dance around the playing field.
Charting cost and mass of each pivot assembly over each year of iteration.
Minimizing these factors allowed for more breathing room on the rest of the bot.
Winch assembly for catapult
Storing the energy for the catapult in advance allowed us to make quick shots once in position.
Deciding on this mechanism was rooted in safety, in the event the robot would shut down the ball would be slowly ejected.
One sim motor in tandem with the transmission offered the torque to stretch the elastic. A pneumatic cylinder was used to decouple the motor from the spool.
The design was given a safety award for creating one of the few bots that had safety as a priority.
Switching to Tank
Each year, the benefits and tradeoffs of the pivot assembly are carefully considered.
The following year, using a tank drive was deemed to be essential for the obstacles introduced in this years game.
Knowing when the simple solution is better is hard to recognize but essential for efficiency.
Bringing CAD Home
Using what I have learned, I have started creating projects with from the comfort of my home using a 3D printer.