Minibot deployment

[philso]

My boys (an award winning FLL team) were asked to build a MiniBot for one of the FRC teams in Houston. We tried several different configurations before coming up with a MiniBot that reliably does the climb in 2.5 seconds. It took a lot of work to install the deployment system on the Hostbot since we could not get any drawings for the it ahead of time. Unfortunately, our Minibot was never deployed successfully in competition.

We found that you need to look at deployment holistically. The Hostbot must be able to get the MiniBot to the pole accurately and reliably. The wobbling of the arm holding the MiniBot while the Hostbot is moving forward makes this more difficult. We also observed instances where the Hostbot drove forward too hard and bounced back, pulling the MiniBot off the pole. The pressure of being in a competitive match may have made a difference in how smooth the robot was driven. A system that does not require the Hostbot to move probably would have worked better.

http://www.youtube.com/watch?v=ibi8ZqC_LXI


A sight (pink pool noodle) was added to the Hostbot to allow the driver to better line up with the pole. This sight only works with the Hostbot in a line between the driver and the pole. Many of the attempts to deploy in competition were unsuccessful since the MiniBot was mis-aligned. One time, it was mis-aligned enough that it was damaged when the Hostbot drove forward and smashed it into the pole.

http://www.youtube.com/watch?v=10MxQCTuAQ0


Also, one has to consider other factors such as shown below. A bit of extra code probably would have prevented this mishap. My boys broke down and cried when they saw this happen, for the second time...

http://www.youtube.com/watch?v=Piq4VSKoy54


We hope others will be able to get some benefit from our experiences.

Phil

[Dad1279]

IMHO the key is in the alignment, and being ready to deploy ahead of time. We used a piece of 8020 with a slide, and a lexan 'V' at the end, so we could align with the pole prior to the 10 second endgame, and be ready to deploy.

Other successful methods included arms that swung out to hit the pole, so they were self aligning.

[notmattlythgoe]

We used a pneumatic scissor system on our robot this year for deployment. This past weekend at Pamletto we successfully deployed 16 of 16 minibots.

[jvriezen]

I may be biased based on what we did, but from what I've seen, for speed and reliability, the best approaches include one or more of the following:

1) Some type of alignment that uses the tower base as a reference (e.g. a fold down plate with a tower base radius matching cutout) This is not essential however, if item (3) is done well. (This is the only item in this list we did not do, but still had pretty good deployment success.)

2) A center mounted slide mechanism, pre-loaded with surgical tubing, and released with pneumatics or servo. We used two drawer slides and one igus slide, with two stages of tubing. Be sure to keep safety in mind -- we put a safety strap (removed just after placing the bot on the field) and also had a red danger flag that we put on the robot when transporting it to help keep people from walking too close behind. For the longer drawer slides, we looped the tubing back and forth to have a greater overall length. In other words, it works better if the tubing goes from 30" to 10" in deployment rather than from 12" to 2".

3) A pivoting V/funnel plate that will help with finer alignment. Ours was integrated with the surgical tubing slider mechanism so that the same laterally mounted surg tubes kept the V centered (until the pole persuaded it otherwise) and also provided slider thrust when released. The plate cutout had a shallow V, but at the vertex of the V, we cut back a rectangular slot, a bit wider than the pole, so that the last bit of travel was straight in line with MB, avoiding an angled approach of the minibot. That's the 'funnel' part of the plate.

4) Minibot mounted with tube/pin interface (at least 2). Probably doesn't matter which bot (HB or MB) has tube and which has pins.

5) Locking pin to prevent MB movement prior to deploy. We used a horizontal pin that went thru a hole in the mounting tube and vertical pin which held MB in place. It had some friction fit, so that it wouldn't jostle loose. The horizontal pin was attached by a string so that the pin would be pulled out just prior to MB reaching the pole. Just gather up the slack in the string with a loosely fastened zip tie so that the string doesn't get caught in something during the match.

5) Turn on power to MB before it reaches pole -- we used a simple string with a loop on the end to pull a standard wall switch on the MB.

6) Software control that releases the MB only when two things are true: a) Driver is pressing deploy button b) Match clock reads <= 10 seconds. Some have told us (b) was risky, as the internal cRio match clock might not be synced exactly with field clock, but we were never called for early deploy.
You mileage may vary. This way we lined up, held the deploy button and as soon as clock hit 10, it goes.


-----
Other good ideas involve a two step deployment, where alignment and contact with the pole happens prior to 10sec, and then the MB rides a rail out the pole at 10sec mark. But this seems more complex and takes more time if you are late to the tower due to being defended.


John Vriezen
Team 2530 "Inconceivable"
Mentor, Drive Coach, Inspector

[Mr V]

Our deployment system is similar to the concept above.

We initially thought about a plate that dropped down or extended that matched the base profile. We quickly threw that out in favor of "flippers" that are deployed pneumatically. On the end of each flipper there is a sw that puts it in an auto mode. If 1 sw is "true" it drives that side at 20% and the other at 40% until both switches are "true", once both sw are "true" it continues to drive both sides at 20%. So if someone tries to push the bot off the pole before the 10 sec mark it continues to stay glued to the pole. This allows the bot to line up with the base within +/- 1/4" of the pole center.

For the actual deployment system we initially tried numerous scissor mechanisms but none had the performance we desired. What we ended up with was some 80/20 on slides powered by doubled back surgi tube. The tray has a "tongue" that fits into a slot in a square tube carrying another cylinder with a piece of round UHMU threaded over the end. The tongue extends past the square tube and has another hole for a safety pin with a long "remove before flight" tag. That is used to lock it back into to position for removing from the field and safety for transport to the field. The surgi tube is connected to the back of the 80/20, the elevator tower, and loops around rollers attached to the square tube holding the retention cylinder. That gives tension even when fully extended giving a second chance if it doesn't attach to the pole on the first try.

The tray also has a "V" shaped slot to account for the +/- 1/4" possible misalignment. The tray has slots the "feet" of the minibot slide in and towers that support the "wings" on either side of the minibot.

Other teams figured out with a big enough slam they could knock it loose. So we added a hole in one of the "feet" underneath the tray. We added a bolt with the head cut off so the shoulder of the bolt passed through that hole. The other end of the bolt was attached to the square tube directly behind it. That locks the minibot on the tray until it is extended. Once that was added to our and one of the teams borrowing our minibot no matter how hard they slammed us the minibot wouldn't budge until it was deployed.

[jvriezen]

Originally Posted by Mr V

Our deployment system is similar to the concept above.

We initially thought about a plate that dropped down or extended that matched the base profile. We quickly threw that out in favor of "flippers" that are deployed pneumatically. On the end of each flipper there is a sw that puts it in an auto mode. If 1 sw is "true" it drives that side at 20% and the other at 40% until both switches are "true", once both sw are "true" it continues to drive both sides at 20%. So if someone tries to push the bot off the pole before the 10 sec mark it continues to stay glued to the pole. This allows the bot to line up with the base within +/- 1/4" of the pole center.

This sounds very effective.. I like the interconnect between the switches and the drive power. I had a similar idea (never implemented) where the switches would simply cause driver control panel lights to illuminate. In theory, you shouldn't get bumped (penalty) after you are touching the tower, but that doesn't stop it from happening, of course.

John Vriezen
Team 2530 "Inconceivable"
Mentor, Drive coach, Inspector

[Mr V]

Originally Posted by jvriezen

This sounds very effective.. I like the interconnect between the switches and the drive power. I had a similar idea (never implemented) where the switches would simply cause driver control panel lights to illuminate. In theory, you shouldn't get bumped (penalty) after you are touching the tower, but that doesn't stop it from happening, of course.

John Vriezen
Team 2530 "Inconceivable"
Mentor, Drive coach, Inspector

The "safe" period doesn't start until the 10 sec mark and we decided to plan to head there early in case there was defense along the way or there were tubes we needed to push away from the base. In later qualifying matches one of the teams we loaned a minibot to in Seattle got knocked out of alignment at about the 12sec mark and they lost contact before the 10 sec mark and were pushed far enough away and pinned just long enough to prevent them from lining up again since the had no alignment system and had to eyeball it. We are extremely happy with how it has worked.

edit: I should add we were happy with how it worked once it was implemented as designed and tested at home. In Seattle we weren't very successful and didn't have as high a success rate as those we loaned minibots to who had no alignment system. That was due to 2 reasons. #1 was the driver didn't deploy the flippers until the bot was in contact with the base, this allowed a greater misalignment than the "V" in the tray could compensate for. The other bigger issue was our lead programmer has a bad case of NIH (Not Invented Here) syndrome. So he threw out the code developed by the align/deploy team's programmer. The original code left the flippers out and the motors going until a separate button was pushed to deactivate it. To make it "better" he combined the switch to cut motor power, retract the flippers, and deploy the minibot at the same time. Since the bot was pushing hard against the base the bumper was compressed and as soon as the motors were shut down the bumper acted as a spring pushing the bot back away from the base resulting in mis-alignment and/or being too far from the pole for the minibot to trip its' latch and lock on to the pole.

[Mpengelly]

Our mini bot deployment works very similarly to others posted on here, surgical tubing to launch the thing out there. Ours however was designed with two things in mind: first, that tolerance was of utmost importance, and second, that having it deploy out the back could cause problems when trying to line up if the towers were surrounded by tubes. At competition, we saw frequently that teams would not be able to align because human players had landed 5 or 6 tubes around the tower and when they tried to back up it just smushed the tube between the bot and the tower.

Here is a video of our in action (at 2:00) and it the same 80/20 bar with sliders but its tensioned forward so you can simply drive up the pole and further and know you will always be lined up:
http://www.youtube.com/watch?v=83OX6P4SSzw

[waialua359]

I have yet to see a team align the way we do. Its not the best way, nor do I really prefer it. However, when our robot weighed in at 119.6 during the last week of the season, we didnt have much choice.
A simple software aid and use of the camera. It has to be dead on, but it does it every single time.

At NYC, it didnt launch for two matches. 1st, we didnt get there in time. Stupid mistake. 2nd, we lost com and never got there. FTA fixed the problem, but we didnt replay the match.
In Hawaii, we played in 17 matches. It launched successfully in 16 of them, except the last one due to a surgical tubing issue that finally gave out after two tournaments. Problem solved for Vegas in a few days.

[jvriezen]

Originally Posted by waialua359

I have yet to see a team align the way we do.

So can you describe how you do it with .4 lbs? Or perhaps a video link? I'm imagining a minibot as the ammo in a surgical tubing sling shot and you just shoot it at the pole...

John Vriezen
Team 2530 "Inconceivable"
Mentor, Drive coach, Inspector

[Bob Steele]

In our second regional we added a camera located directly behind the minibot deployment. this made a huge difference and we never missed the pole once. Being able to sight right down the deployment made lining it up very easy. We only had to be in contact with the tower and then look through the camera...

If you have the weight... I would highly suggest this strategy to any team.

You don't need anything special.. you just need to put the camera in line with the line of deployment ... if you want to get fancy you could put little tabs on your minibot... then sight right down the line and center the pole between the tabs...

This is very similar to using a scope on a rifle...

I am sure you will like it if you try it...

Good luck on the field!!

[Alex2614]

http://www.youtube.com/user/JakeLenh.../5/17HhD-_o_7w

An example of two different methods. For some reason, 180 backed away from the tower (I don't know why, their minibot always worked really well). But then we gave it a shot. Ours is a tray with a hole that the minibot's shaft sits in, and theirs is a swinging method.

[EricS-Team180]

[quote=Alex2614;1046780
An example of two different methods. For some reason, 180 backed away from the tower (I don't know why, their minibot always worked really well). But then we gave it a shot. Ours is a tray with a hole that the minibot's shaft sits in, and theirs is a swinging method.[/QUOTE]

If I recall correctly, the pin holding the mini-bot jammed in the swing-arm after we released it (pneumatically actuated). We widened the hole, a bit, after that.
Eric

[Businessman41]

We used the kiss rule in designing ours. We had a drawer system, rigged to surgical tubing, so it acted sort of like an underpowered sling shot. A servo, which was rigged to hold it in place, acted as the trigger. Our camera was behind the minibot with a screw sticking out to serve as a sight. In order to deploy, you had to push both the 11 and 6 buttons on the Logitech joystick, which is hard to do accidentally. Our team diligently worked hard at figuring out this system after finding out our first one was illegal due to incorrect plastic. Sadly, it took the whole three days to figure it all out, and we never got to try it. It was ready to use in the finals, though, because we ended up being first backup. Not bad for a team with 1/3 of a robot (our arm was also deemed illegal due to an improper motor.)

[B-Bot]

So far the mini bot deployment system that have worked well, are slides and numatic cylinders. Also a bridge that attatches to the poll, so that during the last ten seconds the mini bot rides from the robot and on to the poll.