DEWBOT VI Build

From DEW Robotics
Revision as of 13:35, 23 February 2010 by Foster (talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

This page contains a day by day breakdown of what we did. Dates hyper-links go to the corresponding photo gallery.

Kickoff

  • After losing most of our team to college, and with only 5 returning members, for 85% of this year’s team FRC Kickoff was a new event. 30 students, parents and mentors drove to Upper Darby High School on Saturday January 9th to watch the events from Dean Kamen’s home in New Hampshire and see a mock-up of this year’s play pieces up close.

Week One (1/10-1/16)

Sunday 10-January - Brainstorming

  • The team met from noon until 4:30 in an intensive brainstorming session. We went through the rules, talked about over all game strategy like "Don't get penalties", and came up with specific strategies for each zone for both Autonomous and Teleop.
  • We then broke into two groups to discuss and brainstorm different drive base systems. We talked about the pros and cons of 4WD, 6WD, Mecanum, Pivot and even a tumbler drive. At the end of the session we were leaning towards 6WD and Pivot drive systems
  • We then split again to discuss shooters and herders along with how to suspend the robot from the towers. We came up with a number of ideas that we will discuss on Monday.

Monday 11-January - Drive base decisions

  • The three hour session was mostly around drive base issues and ideas so we could lock into a drive base decision. To help the discussion we made Paper Dolls of the bumps in the field, 4WD and 6WD with a variety of wheel sizes and a mock up of the tumbler.
  • There was lots of good discussion around what could work well and what the problem areas would be. At the end we decided to move forward with the pivot drive base. Clem is going to look at what the gearing will look like for 6" and 8" wheels.
  • We also got a demo of Google Sketchup a drawing tool that can be used to make quick drawings. We also got to see how SketchyPhysics (a physics plugin for Google Sketchup. It combines a simple and easy to use interface with the power of a fast and accurate physics library) works. It will allow the students to create some interactive (and accurate) models. A small team was started on creating models for:
    Flipper (two flippers that will act like pinball flippers to move the ball.
    Sweeper (Rolling tube with vanes that will roll the ball)
    Kicker (Foot on a leg to kick the ball, along with guide plates)
    Pusher (Flat plate with two pneumatic cylinders)
    Cory Foot - a foot on a rotating arm that extends as it rotates
  • Meanwhile, down in the shop, the construction of the bumps started. Matt, Alex and Faith got things underway by getting the first few sheets of plywood cut. The goal is to have all field components complete by the 19th.
  • Tuesday - No meeting on Tuesday

Wednesday 13-January - Field build and more design

  • More progress on getting the field components cut out and assembled. The bump is looking good, we can start to see how interesting this will make the game

Thursday 14-January - We have a Bump!

  • The bump got completed and the field team started cutting parts for the goal. Kicker prototypes have been drawn in Sketchup/ShetchyPhysics and some ideas will move from the simulation to the prototype process.
  • Friday - Rest for the weary, no meeting

Saturday 16-January - Our First All Day Build

  • Lots of activity with lots of people working is the best description of the day.
  • Programming made great progress. After a kick off meeting the team set about doing a number of different projects.
  • With the robot still in the design phase the Electrical team is not ready to do any major work. They did some work on testing battery systems and doing some minor repairs.
  • The largest amount of progress was made with the Field Components. casters and a large handle were added to the bump to make it super easy to move. The tower supports were completed and a test of some lighter weight pipe showed that we need to use the Schedule 40 pipe (or find lighter roboteers). Progress on the goal continues.
  • There were lots of flying soccer balls with the kicker prototypes in full swing.
  • Thanks to Anne and Rita for feeding us great pizza with garden salads for lunch along with chicken and Mac/Cheese for dinner. The oranges made a great mid-day snack.
DEWBOT VI Week One Build Season Photo Gallery

Week Two (1/17-1/23)

Sunday 17-January - Jumping into Week Two

  • We started week two off with a bang with the sounds of balls hitting the wall. All of the teams made good progress towards their goals.
  • We've got a pretty good kicker prototype. Made the ball fly (but didn't break any clocks).
  • Pivot parts drawings completed for transfer to Machinist.
  • Field Equipment is proceeding well. The goal is very big and voluminous
  • Programming is moving forward with the vision system.
  • Electrical parts have been sorted and reorganized for easy access.
  • Thanks to Andrea for homemade meatball sandwiches and pasta salad.

Monday 18-January - Moving the ball

  • The air spring kicker didn't turn out the way we thought it would
  • But the latex tubing kicker is a trooper
  • It's all about kicker momentum at the time/point of impact
  • The depleted Uranium kicker (Rizzo test) made no difference - heavier kicker is slower to accelerate - hits ball at slower velocity - same momentum - same travel
  • The CAD team got a CAD workout
  • Field equipment getting close to complete

Wednesday 20-January

Thursday 20-January

Saturday 23-January - Cutting the frame

  • Cut almost all frame elements
  • Dry-assembled and tested all steering motors
  • More kicker testing and detailed design
  • Bumpers started
  • The final field element - the tunnel/platform - is finished
DEWBOT VI Week Two Build Season Photo Gallery

Week Three (1/24-1/30)

Sunday 24-January - Welding day

  • We've got a chassis frame (excluding gussets). Matt did a great job welding this in a day.
  • Steering motors epoxy-bonded to steering drive sprockets w/ encoder magnets
  • First magnetic encoder mounted on steering motor and tested. Excellent linearity.
  • All chains broken to right lengths
  • dxf of Pivot Plate Top completed
  • Demonstrated proof-of-concept for line-rider
  • Camera working and acquiring target

Tuesday 26-January

  • Most of Tuesday was spent in a team meeting going over where we are in the design process and making sure our strategy / plans are on track and understood by everyone on the team.

Wednesday 27-January - Full build ahead!

  • Had 16 students show up tonight- Brianna, Amanda, Carly, Christine, Selina, Ben R., Nicole,Sasha, Cole, Andrew, Doug, Alex, DJ, Paul, Preveen and Zain.
  • Jon worked with programmers, David helped Jon with game strategy. Praveen, Sasha, Paul, Ben R., Zain, Nicole. DJ got the in-house chat working but it needs to be put on a server..... Now he will be learning LabVIEW to use it. Paul is making a LabVIEW disc for both DJ and Alex-will be ready tomorrow-Ben R. to help DJ learn LabVIEW.
  • Metal cut for bumpers / screws bought to install them-Faith, Doug, Cole
  • Kicker mounts marked and cut-Selina & Carly
  • Trident built- Julie, Christine, Alex, Amanda Brianna
  • Scott and Garry Andrew, Sasha, Carly started electrical
  • Rizzo and Cole fit and marked gusset positions

Friday 29-January

  • Tuffak Polycarbonate cut for Battery Mount (put in oven for conditioning prior to forming)
  • Tuffak Polycarbonate rough-cut for Jaguar & cRIO electrical panels
  • Compressor mounts and front gussets cut

Saturday 30-January - Drive train push

  • Machined pivot parts received from sponsor Wamac (Thanks!!)
  • Pivot assembly team went into action, directed by Gary Deaver
  • A set of bumpers (the blue ones) were completed (Mike Rizzo leading team)
  • Chassis frame modified to accommodate kicker and gussets welded in (Matt welding with Siri mentoring)
  • Programming math worked out for the Pivot Drive
  • Kicker design approaching completion
DEWBOT VI Week Three Build Season Photo Gallery

Week Four (1/31-2/6)

Sunday 31-January - Drive train Focus

  • The shop was alive with activity around the drive train. Getting the pivot assemblies complete, getting drive chains cut to length, getting the motor mounting plates riveted in place.

Monday 1-February - Last minute chassis items

  • Now that the chassis is complete, things that attach are moving through the construction pipeline. Side panels that hold Jaguar speed controls were cut out and drilled. Same for the rear panel that holds the cRio and digital side car. Tension adjustment holes for the chains were added.

Tuesday 2-February - Is Punxsutawney Phil right, only three weeks to go?

  • Today starts the big push, only 21 days before the robot ships. Tonight four teams worked on different components.
  • Metal for the second set of bumpers (red ones) arrived today. The bumper team got it cut to the proper lengths and drilled the holes for attachment. Their task for Wednesday is to complete the red bumpers. This is the earliest we've ever had bumpers, a new DEWBOT record.
  • Over in Electronics another new record was set. All of the Jaguars and the cRio were mounted and wired in under an hour. Having the panels on the bench rather on the robot helped, but having 4 people work on the wiring helped the most. Everything was cut, crimped, labeled and attached in record time.
  • Welding of the kicker assembly was completed and the frame is now considered to be 100% complete. From this point on it's attaching components to make the robot move and score.
  • Programming team worked on trials and tests with different IR sensors to see how well they can find the balls for autonomous mode.

Wednesday 3-February - Chassis Assembly Night

  • The focus of the team was to get all the pivots, drive and steering motors and electronics mounted on the frame. It was a long build session with lots of components coming together.
  • The control team went home, so Foster fired up a VEX controller to make DEWBOT VI roll. A few lines of C code and we were set to test it out. (Thursday should have it under cRio power)
  • The loud screams you heard at 10:10 PM were from team Sab-bot-age watching DEWBOT VI take its first roll around the shop. The pivot drives work very well, the robot moves forward and back and can strafe side to side. Clem drove it over the bump four times, so it was a major milestone evening!

Thursday 4-February - Kicker, Bumper, Cleanup and cRio

  • Tonights activity focused on four areas, Kicker, Bumper, Cleanup and cRio.
  • Ken and Foster spent most of the night working on the robot trying to get the electronics running. All of the wiring is complete (needs zip tied) but is reading for testing. Ken struggled with getting all the components to talk to each other. Some subsystems like the camera work, others like the drive train does not.
  • Mark, Cole and a few others worked on getting a duplicate frame made for the kicker. The idea is that the electronics and drive teams can work on the drive base getting it to work and for driver practice. At the same time the kicker team can use the duplicate base to improve the design. It's a little extra work, but it will be worth it in the long run.
  • Rizzo and his team did some minor modifications on his bumper design. The blue set is complete, the red set has all the materials here. The metal angles for the edges has been cut, we need a few more feet of pool noodle material to start assembly.
  • Matt and Faith lead the teams to do the mid-season shop clean. With the large amount of materials that comes in there is a ton of boxes, wrapping, etc. They put tools away, threw trash out, and did a complete reorganize. The shop looks much better, is easier to find things and looks a lot less crowded.
  • Hopefully we will close out week four with a working kicker and a robot that really moves.

Friday 5-February - Who the heck scheduled a blizzard during build season?

  • Blizzard? What blizzard? The snow didn't start until after 5PM so we're here anyway.
  • We've become programming-centric. Mechanical & Electrical teams are supporting the Programming team. Nicole, Ben and Ken worked on LabVIEW code for the drive train. Jon was able to fabricate a mounting for the motion sensors while Scott, Nicole and Carly wired power to the daughter cards on the cRio. (And Carly did some shop electrical repairs).
  • To facilitate workflow, we are building a proxy chassis to allow us to build and further test the kicker (and develop centering & possession systems) without taking the chassis from the programming team (and subsequently, the field team). Clem McKown churned out part and assembly drawings while Faith McKown, Matt & Gary Deaver cut metal. Foster cleaned the shop in preparation for welding. We'll weld tomorrow if we can get in (not quite all of us got home!).

Saturday 6-February - The Snowpocalypse!

  • Downingtown and the surrounding area got 16-18" of snow over night. So the shop was closed and we will start Week 6 on Sunday.
DEWBOT VI Week Four Build Season Photo Gallery

Week Five (2/7-2/13)

Sunday 7-February

  • Still need to get the robot moving. Not there yet. No pressure.
  • Robot weighed in at 89.0 lb (without kicker, Clippards,...). Weight will not be a problem.
  • Based on feedback from Chief Delphi, we decided to fix our bumper gap problem (Mind the Gap!). Mission completed. We had to mine pool noodles from old bumpers (have you ever tried to find pool noodles in February?).
  • Welded most of the proxy chassis before the MIG gave up the ghost. We'll pick up a new feed liner Monday.
  • Gary Deaver's got a good handle on the possessor (we think). No not really.
  • Metal is cut for the kicker. Need the MIG working for this too.

Monday 8-February

  • We repaired the MIG welder, another example of how a $7 part can make all the difference in the world.
  • Completed the proxy chassis
  • Bumpers are finished - red & blue
  • Partially assembled the crate and checked that we have all the hardware
  • We also had a visit by Dr. Mussoline, Superintendent of Schools, and Dr. Watson, Assistant Superintendent of the Downingtown Area School District (DASD). They toured the shop, got a quick overview of Breakaway and a chance to see the robot. DASD is one of the teams largest sponsors.

Tuesday 9-February

  • Welded the kicker this evening. The quality of our welding has improved significantly. We are moving from making strong welds to making (strong &) beautiful welds. not quite there yet, but we've turned a corner.
  • We're mounting systems on the proxy chassis. The kicker is on (cocking arm not yet). Gary Deaver's got a prototype possessor mounted which needs some testing.
  • Gary Deaver successfully mounted a Fisher Price (-9015) motor to Bainbots P60 planetary gearbox.
  • Ordered 3-stage ring gears from Bainbots (to speed things up)
  • Programming team made progress this evening. Pivots more stable and making "steering-like" movements
  • Rizzo & Jen are working on a giant screwdriver (our logo). Gotta see this!
  • Left a little early because the roads were getting bad. We are hunkering down for another big snowstorm. (School &) Robotics are already canceled for tomorrow.

Wednesday 10-February

  • Another huge snowstorm - no meeting!
  • Work was done on the website
  • And some design as well
  • Gary Deaver posted the Pivot photo on Chief Delphi

Thursday 11-February

  • Integration Meeting - final decision on sensors and what last purchases
  • The kicker is on the proxy chassis. Some modifications will be needed.
  • Pivot movement is somewhat more steering-like
  • Battery, Power Distribution, Camera & Disconnects positions revised - final mounting tomorrow

Friday 12-February - a Huge day!

  • The suits came in early to solve the drive programming issues
  • The robot moves under cRIO control!!!!
  • Weigh-in - DEWBOT VI weighs 89.6 lb without kicker, Clippard air cylinders & a few other items
  • Stripped all electronics off the robot for final mounting work
  • Drilled holes
  • Tapped holes
  • Made up mounting plates
  • Ran a successful test of the possessor
  • Quick welds made to correct a Cocking Cylinder issue (on both the robot & the proxy chassis
  • And to add the Beacon's mounting bracket
  • We mined DEWBOT IV for 80/20 parts
  • And thought a lot about the centeror
  • Then we reinstalled & rewired all of the electronics
  • Turned it on - it works!
  • Thank God for snow days!

Saturday 13-February - a really good robot day

  • We've got Crab mode working really reliably
  • Went over the bump time & again. No tips (robot is really stable). Lost the battery once in a hard dismount. We'll need to add a Velcro strap to secure this.
  • We got the kicker working on the proxy chassis. Elastics run. Pneumatics installed. Good kicking. Got some tweaking to do to optimize the kicker.
  • The possessor is mounted (proxy chassis) and working
  • Much of the kicker release was fabricated today.
  • Started on the pneumatic equipment mounts for the robot (Clippard, regulator, etc.)
  • An elegant, simplified camera mount was devised. Need to convert to polymer.
  • Have brought many of the recent mods into the Inventor model.
  • The Rizzo screwdriver is taking shape
  • Bumper weight check - combined mass of both red & blue bumper sets is 25.2 lb
DEWBOT VI Week Five Build Season Photo Gallery

Week Six (2/14-2/20)

Sunday, February 14 - Valentine's Day

  • A day for driving practice & refining the drive control code
  • Kicker modified and performing well. An assist cocking cylinder was added and is necessary to drive the elastics
  • Kicker release installed - some additional development needed
  • Started mounting pneumatics (beyond the compressor) onto the robot
  • Confirmed that the robot can easily drive over balls sideways. Need to address this.
  • Painted the inner wooden faces of our bumpers (red & blue, naturally)
  • Students worked on Chairman's Award submission
  • Screwdriver taking shape in a big way

Monday, February 15 - Presidents' Day

  • We learned to watch battery voltage very carefully. Jaguars are sensitive to low batteries.
  • It's really hard to herd in Crab
  • A rudimentary Snake mode is working
  • IR sensors mounted and tested
  • Pneumatic system is on-board (with surrogate cylinders for kicker & release). Compressor code tested.
  • Still working on the kicker release
  • Driving and herding experience is showing that we will have troubles with driving over balls in the robot front with the possessor (unless we slow or reverse it - can this run on a sensor?)
  • Mark Morun created a chain-guard form (for vacuforming) that is a work of art

Tuesday, February 16] - 2 Breakthroughs

  • Read the punchlist
  • Breakthrough - Snake drive mode is now working (and working well). Uses the mathematics developed over the summer and applies a square root function to avoid oversensitivity with small joystick displacements
  • Breakthrough - The kicker & release system seem to be working reliably. David Moyer, Mike Rizzo, Douglas and Clem McKown stayed late to work on the problem. The issue was that at the very high contact pressures, the release system wasn't releasing reliably. When it did release, it rapidly wore out the steel contact surfaces (after 10-25 kicks). A tool steel contact plate reduced wear, but did not provide reliability. Lubricants provided good operation, but only for a few kicks. The breakthrough involved placing a couple of wire-ties on the contact surface so that the contact was between the relatively lubricous wire-tie and the tool steel. This worked for more 21 kicks without showing signs of wear. A great job!
  • Gary Deaver also assembled an alternative release mechanism based on a hose connector. Very cool.
  • Chain guards - Tuffak polycarbonate pieces are cut and conditioning in the oven prior to vacuum-forming
  • Robot Top - Another piece of polycarbonate was cut for the robot top (to keep balls out). Detailed cutting will take place tomorrow.
  • Side guards - Mounting holes drilled & tapped. Eye bolts will be in tomorrow.
  • Battery strap made
  • Team numbers painted onto the red bumper set

Wednesday, February 17 - Crunch Time

  • Read the punch list
  • The kicker and release system started chewing up wire-ties today. Rizzo did an excellent root-cause analysis, realizing that if the kicker is not pushed all the way back, the latch sheers the wire ties. Need a limit switch here. We were able to return to David Moyer's original 1-cylinder latch arrangement. We also returned to reliable kicking. We will mount the existing kicker on DEWBOT VI (tomorrow), but also fabricate a spare.
  • Side guards complete and work perfectly. We cannot drive over balls from the side with these in place and can herd balls safely with the robot's sides. They do not interfere with crossing over the bump.
  • The Possessor is mechanically and (mostly) electrically installed.
  • Cut polycarbonate and a form for a passive Centeror.
  • Detailed design of robot top is almost complete.
  • Revised the Sponsor page, adding logos & links.
  • Completed the slide show.
  • Weight check: we're at 118.0 lb, without chain guards. Ideas: 1) use thinner polycarbonate for robot top & chain guards (3/16" currently); 2) reduce polycarbonate panel sizes of underutilized panels; 3) reduce lower flange on kicker face (non-contact); and 4) use plastic hubs in lieu of steel for the kicker axle. There's > 5 lb opportunity here without compromising performance.

Thursday, February 18 - Under Pressure!

  • Read the punch list
  • Re-enforced the cocking cylinder mount - these welds failed on the proxy chassis.
  • Worked on installing the Kicker on DEWBOT, but didn't finish the job.
  • Submitted Chairmans Award docs
  • Arranged to obtain 1/16" polycarbonate for robot top fast.
  • Paul wrote kicker code
  • Wired the Possessor.
  • Labeled wires
  • Completed robot top design
  • Completed shirt design
  • Installed Battery Strap
  • Adjusted chains & sprockets
  • Applied Locktite to bolts

Friday, February 19 - Tick Tock!

  • Read the punchlist
  • The Kicker was successfully installed onto DEWBOT VI. Gary Deaver had the idea of using only the 3/4" helper cylinder to draw back the cocking arm (the 2" cylinder only pushes to cock the Kicker). This works and significantly reduces air usage, reduces battery draw and should allow more (and more frequent) kicking.
  • The 1/16" polycarbonate arrived for the robot. Started cutting this.
  • Cut metal for a spare kicker
  • Mark Moran and Cole worked on the Camera mount, which is nearly complete.
  • Ben R worked on autonomous code.
  • Douglas finalized the T-Shirt design, which was submitted to the printer.
  • Slide show under construction.

Saturday, February 20 - Down to the wire!

  • Read the punch list
  • Formed & installed the passive centerors
  • Installed & wired the field-position switch
  • Painted "1640" on the last two bumpers
  • Cut the robot top
DEWBOT VI Week Six Build Season Photo Gallery

Final Few Build Daze (2/21-2/23)

Sunday, February 21

  • Big wire-tie party! Wiring is much more organized now
  • Finished & installed the robot's top
  • Drive code working (again). Kicker code working.
  • We got in some quality driving & kicking practice.
  • Learned that the Possessor really doesn't work. A revised version also failed to perform. After a series of tests, we settled on a passive, low-impact ball positioner in front of the kicker.
  • Formed, finished & installed 2 more chain guards (1 to go!)
  • We can go through the tunnel
... and over the bump

Monday, February 22

  • It was a basic nite to clean up stuff.
  • Started with replacing a very low battery.
  • Tuffak top had remaining cutouts color coded. Black for lifting, white for control, red for danges areas. We used tape, we are searching for some other material.
  • Andrew put the last of our 2006 sponsors on-line so we have a complete list for all the years.
  • Space that was used by the crate got filled with the game parts
  • Programmers had a issue with the code. Turned out to be a problem with the digital sidecar, the cable to the cRio was not connected 100%. Another item to add to the pit check list.
  • Ben was able to make the simple auto version run. So we have a working autonomous. (yay)
  • The last chain guard was formed and fitted to DEWBOT. Safety and beauty at the same time.
  • New wheel treads were installed all around
  • Most of the rest of the evening was building the crate and securely crating DEWBOT VI for the trip to RIT. We brought the crate out for pick-up a bit after 1:30 AM.
  • Rizzo with some help from Carly and Foster built a new cart to carry the robot on and off the field. It was not perfect, but does a great job of holding the robot upside down which makes it easer to work on the robot. We will refine the plan and build #2 this weekend.
  • Final DEWBOT VI mass is 112.2 lb (versus 120 lb max). Bumpers weigh 12.6 lb per set (versus 20 lb max).
  • Center of Mass (measured) is 17.5" from the frame rear (which is dead center between the front & rear wheels; exactly where we wanted it) and 9.75" from the floor.
  • Screwdriver is about 75% complete. It currently weighs 17 lb.
DEWBOT VI Final Daze Build Season Photo Gallery