PDA

View Full Version : Outline- Preparation for Competition Readiness



DresnerRobotics
07-03-2008, 09:36 PM
Year 1 is going to be a year of learning and experimentation. My only real goal is to get out there, make it happen, and have fun.

This is mostly aimed at the newcomers, people who have been following the discussion and birth of this idea probably already know what is involved all too well. Just something to keep in mind for anyone planning on participating, and I don't mean this as a discouragement but rather just a reality check-

This competition has probably one of the highest bars in terms of technology and skill required to compete. I'd say Robomagellan is an example of something more difficult, there aren't many. There is a lot involved in making a bot capable of competing in Mech Wars.

Just as a rough outline, here are the challenges you face:

Walking- this alone is not an easy task. Biped walking gaits that are stable and mobile are things that even university research labs are still working to perfect.

Balance- keep in mind if(when) your bot falls over, that will count as a damage point against you. You also need to be able to right yourself otherwise you're looking at another penalty point. You should definitely look into the use of IMUs.

Vision- you need to think about implementing a wireless camera system. That said- we are looking into options for an easy to use, small wifi color camera. These units arent normally available retail but we may have a special connection for them, and will be making them available at our cost.

Target/scoring system- We're making this standardized, and taking the load off of the builders. We are in the process of developing a standardized, wireless, modular target/scoring slave units to provide to competitors at cost (est $50 USD) You will still need to mount micro-switches on target plates to your mech (standards for those yet to be decided), but they will simply plug into these scoring slave units and everything else will be taken care of.

Onboard controller- You'll need to decide what onboard brains you'll be using. Your bot isnt going to walk on its own.

Wireless comm and control setup- you need to come up with how you plan on controlling the walking gaits, aiming, and weapons. There are many options out there, Xbee, bluetooth, wifi, various RF. Also need to decide what you plan on using for a remote controller.

Programming- The glue that will tie this project together. You will very likely have to have a solid understanding of a programming language versatile enough to incorporate your remote controller, onboard controller, walking gaits and weapons systems. I am working towards a solution using the Bioloid system, and will make it available in detail to the public when I am finished, but do not bank on this being your only option.

Not trying to discourage anyone here, just trying to outline what is involved in this project. Good news is we have just under a year until our first event.

darkback2
07-03-2008, 10:08 PM
Just a thought, and I think I may have recommended this before. One way to eliminate the problems with frequency and what not, why not get 4 video camera/transmitter/receiver setups and make them available at the event. In that way you don't have to worry about people interfering with each other. Everyone would just have to have a place to mount your camera on their robot. Same thing with the targeting system. It would be better in my opinion to have targets that mount on everyones robot, and have the targets transmit to a central computer that would keep score. The target system could have an output for people to connect to say a phidget 8/8/8 or other sensor input if they want to integrate it into their robot programming.

I hope you don't take any of these suggestions the wrong way. They are just ideas.

DB

Adrenalynn
07-03-2008, 10:26 PM
A few concerns I'd have with that suggestion, maybe you can explain it more?

How would they integrate and test it without having it ahead of time?

Where does the money for Tyb to absorb all these costs come from?

How does "one size fit all" ?

darkback2
07-03-2008, 10:45 PM
How would they integrate and test it without having it ahead of time?
The video system? If everyone had a mount somewhere on their robot that could fit a camera system of a given size and shape...I'm assuming that the four camera's would be the same?" You could use any camera you want at home...It's more about avoiding conflicts at the event. People have been posting about the limitations of various sytems...only so many channels and so forth. I'm just looking for a way to make it so that there aren't any problems.


Where does the money for Tyb to absorb all these costs come from?
Everyone who is participating agrees ahead of time to pay a participation fee seperate from the fee's associated with Robogames. For example. A camera rental fee...target rental fee. Actually though, if the targets are $50 then that would be ok for everyone to buy their own...I guess I'm just thinking of ways to make this game work as smoothly as possible. Also, I'll put my money where my mouth is. I'll make a contribution. We are going to need to anyway because of the cost of the arena, buildings...There are a lot of costs involved with putting an event like this together. I figured we would all be paying...just it would suck to pay for something and come to find out it conflicts with a competitors thing...



How does "one size fit all" ?
It doesn't...more all fit one size.

Again, I'm not trying to nitpick or whine...I'm just throwing out ideas. Hope this helps.

DB

DresnerRobotics
07-03-2008, 10:58 PM
Just a thought, and I think I may have recommended this before. One way to eliminate the problems with frequency and what not, why not get 4 video camera/transmitter/receiver setups and make them available at the event. In that way you don't have to worry about people interfering with each other. Everyone would just have to have a place to mount your camera on their robot.

Definitely a possibility... I think we should see how Adrenalynn's wifi camera source turns out first before we look elsewhere. That would be the best solution as all the 4 channel camera setups Ive found are prone to interference, as they are 2.4ghz.



Same thing with the targeting system. It would be better in my opinion to have targets that mount on everyones robot, and have the targets transmit to a central computer that would keep score. The target system could have an output for people to connect to say a phidget 8/8/8 or other sensor input if they want to integrate it into their robot programming.Check out this thread :) http://forums.trossenrobotics.com/showthread.php?t=1931



I hope you don't take any of these suggestions the wrong way. They are just ideas.

DBNot at all! I appreciate the feedback and suggestions. Thank you.

I would however just as soon avoid any 'registration/participation' costs unless they become absolutely necessary. We'll see how things go after year 1, if I'm having to pay tons out of my own pocket to make it happen then it's something I'd consider... but it would be explicit as to what it would be going to. For now, I think people have enough cost to cover getting their mechs put together, I'd rather not add to that.

dcalkins
07-04-2008, 02:24 AM
The video system? If everyone had a mount somewhere on their robot that could fit a camera system of a given size and shape...I'm assuming that the four camera's would be the same?" You could use any camera you want at home...It's more about avoiding conflicts at the event. People have been posting about the limitations of various sytems...only so many channels and so forth. I'm just looking for a way to make it so that there aren't any problems.

Great idea in theory. In practice, mounting options may be extreme from bot A to bot B. Also, some users may hard-wire a camera into their bot, behind armor, for example. Finally, some builders may want 2 or more cameras on their bot...

Frequency control is ALWAYS an issue at these events. It's my single greatest concern. I deal with it more than you [sincerely] could possibly imagine. R2-D2 and hockey bot @4 on the same freq. WiFi hubs next to Specrum transmitters. Idiots with uncharged battery packs insisting someone is jamming them...

For 1:1 mech, experience tells me that it won't be an issue at this range. All 2.4G cameras I know offer 4 channels. Although it should be clear in the rules (and the safety check-in proceedure) taht the user must demonsrate:

1) multiple channels on the camera. Even 2 will suffice. So you must be able to broadcast on 2 different channels or freqs.

2) Unique Tx/Rx sommunication between controller and robot. [stuff builders don't have to deal with but I do:] Let's say you're testing your bot. Tx is off, but robot is on and loaded. Some innocent guy 100 feet away is playing with his RoboOne, which is the same freq as your mech. You get shot in the eyeball. Who's fault? Not his, not yours. In theory. In practice, it's the rules-author. All mech's must have unique Tx/Rx pairs, such that no Mech transmitter may operate any other robot, and no other transmitter may operate your mech. This will be written into the rules, and validated at check-in. Just like the ComBots.

Stickler on safety, I am.


Everyone who is participating agrees ahead of time to pay a participation fee seperate from the fee's associated with Robogames.

FYI: The $35 fee is about equal to what it costs us for a bot. An 8x2 table with 2 chairs for the weekend is $22. BuildersDB fees per robot is $2. Electricity is by kWh, and we figure it's about $3/bot/weekend (those chargers of yours suck watts..) Badges are about $2/person. Envelopes and builder packets are around $1/team. etc.


For example. A camera rental fee...target rental fee. Actually though, if the targets are $50 then that would be ok for everyone to buy their own...I guess I'm just thinking of ways to make this game work as smoothly as possible.

My general take - rentals don't work. Just like you gotta buy a charger, or a control transmitter , you gotta buy a camera. And anyone who doesn't have 20 hours of driver training before arriving is a FOOL. Tele-operation is hard. I have a mock-up lunar rover, and I tell ya... After 40+ hours I can still barely drive it by watching the screen. And thats a four wheel platform.

And targeting system? Builders need to have it attached before they arrive. No BS about "I can't get my targeting system attached" delaying a match for 2 hours...

No... builders need to buy their own camera and targeting system before arrival. And USE it before arrival. For their own sakes....


Also, I'll put my money where my mouth is. I'll make a contribution. We are going to need to anyway because of the cost of the arena, buildings...There are a lot of costs involved with putting an event like this together. I figured we would all be paying...just it would suck to pay for something and come to find out it conflicts with a competitors thing...

No... I'll absorb playfield costs. I think that it's unfair for builders to expect Event ORganizers to absorb robot costs (e.g., cameras) and I think it's equally unfair for event rganizes to expect builders to absorb playfield costs.


It doesn't...more all fit one size.

This is also on me and Andrew. We need to spec the targeting system, and let builders know that (for example), "targets are 4 inches wide by 3 inches tall by .75 inches thick and weigh 200 grams. They have a built in 3 V regulator and can take 3-12VDC input, via an "m" class plug." or whatever.


Again, I'm not trying to nitpick or whine...I'm just throwing out ideas. Hope this helps.

No one thinks you're whining. Thanks for the input.

DresnerRobotics
07-04-2008, 03:04 AM
This is also on me and Andrew. We need to spec the targeting system, and let builders know that (for example), "targets are 4 inches wide by 3 inches tall by .75 inches thick and weigh 200 grams. They have a built in 3 V regulator and can take 3-12VDC input, via an "m" class plug." or whatever.


Yup, Jon is in France this next week but I believe the RF chips he ordered for the slave/master target/scoring units should be arriving soon. The three of us will have to pow-wow on this. Somehow I think the decision on target plate size standard is going to be one of the tougher ones... might need to have some flexibility on it (for example, my bot doesnt have a big flat chest, dunno how Im going to work that one out).

Eh, we'll get something figured out.

Electricity
07-05-2008, 12:52 PM
This is also on me and Andrew. We need to spec the targeting system, and let builders know that (for example), "targets are 4 inches wide by 3 inches tall by .75 inches thick and weigh 200 grams. They have a built in 3 V regulator and can take 3-12VDC input, via an "m" class plug." or whatever.


I know your just throwing out random numbers, but might I suggest the use of an accessory battery pack for the targets? This would eliminate the temptation for builders to incorporate some sort of hidden kill switch that would prevent the target from getting power, therefore not registering kills..

DresnerRobotics
07-05-2008, 01:26 PM
I don't think it will be necessary, we'll validate the scoring system is working before we start a match, but who knows- we'll see how the design goes.

4mem8
07-05-2008, 08:24 PM
Do you think in the future [Not now or year 1] that when you receive a direct hit [top score] for that hit plate that the Mech automatically stops and a smoke cannister of some kind gets activated for realism, and the Mech powers down, Neat effect.

DresnerRobotics
07-05-2008, 09:59 PM
Do you think in the future [Not now or year 1] that when you receive a direct hit [top score] for that hit plate that the Mech automatically stops and a smoke cannister of some kind gets activated for realism, and the Mech powers down, Neat effect.

Absolutely. This was something Fon and I discussed to add a degree of realistic 'battle damage'. Not something I'll bank on at year 1, and we need to be careful what we're emitting into the air inside Robogames, but certainly something that should be considered for the future.

4mem8
07-06-2008, 01:43 AM
Cool, I look forward to that implementation in time.

MSpurk
07-12-2008, 09:23 PM
Just to clear the air I hate rules, but it is probably necessary to add a rule about no shields or intentionally blocking a target plate. I've never played with Airsoft guns, but hitting a 3x5 index card is tough with a NERF gun. hitting it with a robot controlled through a camera is going to be brutal. Plus, I doubt anyone's bot is going to be large enough to accomodate a 3x5 target surface on both sides. That is huge on these small bots.

PS I already have my quad walking. I'm going to be rebuilding the entire frame and legs to make them look cooler. I want to get the walking gait better then we can start on the turrets/armor/guns.

-Matt-

DresnerRobotics
07-12-2008, 09:37 PM
Just to clear the air I hate rules, but it is probably necessary to add a rule about no shields or intentionally blocking a target plate. I've never played with Airsoft guns, but hitting a 3x5 index card is tough with a NERF gun. hitting it with a robot controlled through a camera is going to be brutal. Plus, I doubt anyone's bot is going to be large enough to accomodate a 3x5 target surface on both sides. That is huge on these small bots.

PS I already have my quad walking. I'm going to be rebuilding the entire frame and legs to make them look cooler. I want to get the walking gait better then we can start on the turrets/armor/guns.

-Matt-

There is a rule in regards to no shields being allowed already posted. I would hope that people participating in this are in it to have fun, and would refrain from cheap tactics just to win. Its not about winning, especially not in year 1, its about getting out there and making this happen and having a blast in the process.

Would love to see some pictures of your quad, why dont you make a thread regarding your mech project and keep it updated as you go along?

4mem8
07-12-2008, 10:40 PM
Definatey Mspurk: Pics please or as Tyb states make a thread. We all want to see it.

majortom1001
08-27-2008, 01:07 AM
Hi All
This is my first post, just joined. I'm interested in maybe competing in this challange. I like that the weight has been increased, in the unrestricted class you might raise the weight for increased weapon load. One other suggestion is not being able to have a overview of the battle field navigating by camera alone might take a long time before the MECH's bump into each other, bad for the viewing crowd. One possible answer might be a local GPS system like the North Star system, at least you will know where you are. Anyways just some suggestions, not trying to come in late and rain on anyone's competition rules. Will try to compete in event what ever final rules are, think this is great and will have a lot of fun.
Thanks

DresnerRobotics
08-27-2008, 05:30 PM
Hi All
This is my first post, just joined. I'm interested in maybe competing in this challange. I like that the weight has been increased, in the unrestricted class you might raise the weight for increased weapon load. One other suggestion is not being able to have a overview of the battle field navigating by camera alone might take a long time before the MECH's bump into each other, bad for the viewing crowd. One possible answer might be a local GPS system like the North Star system, at least you will know where you are. Anyways just some suggestions, not trying to come in late and rain on anyone's competition rules. Will try to compete in event what ever final rules are, think this is great and will have a lot of fun.
Thanks

Welcome! Glad to have you onboard. With a 5kg weight limit, I highly doubt anyone will run into building restrictions. Even the heaviest Robo-One bots are around 4kg. It's also there for safety reasons.

I think the arena will be small enough that it shouldn't be an issue finding your enemy. We can always adjust things on the fly at the competition to make things work better. This first year is sort of a wait and see, learning year. GPS would be cool but unfortunately I doubt it would be very accurate inside Robogames, and payload is a very precious resource in building a mech.

MSpurk
08-27-2008, 07:37 PM
If we find the Mech's are having trouble finding each other we can always have the Mech "pilots" designate a spotter to help them determine where the enemy Mech is located. It will eliminate any sneak attacks, but in the event it starts getting dull it could help. It will be similar to a radar operator calling out enemy units.

-Matt-

Adrenalynn
08-27-2008, 07:47 PM
If it gets too boring, we could just toss my Jack Russell Terminator into the ring... He won't have any trouble finding and disposing of them. :robotsurprised:

MSpurk
08-27-2008, 08:04 PM
Off-topic alert

Oh, you have a terrier....I'm sorry. I hope you were at least paid well when it was forced on you. :wink::veryhappy:

-Matt-

Adrenalynn
08-27-2008, 08:13 PM
Shhh! He knows how to use the GPS. He has your IP address. If he sees that, you are so getting your leg torn off and it kinda sucks being beaten with your own bloody leg-stump, if ya catch my drift. ;)

Electricity
08-28-2008, 10:57 AM
lol.


I wanted to say more, but thats all I could think of..

DresnerRobotics
11-04-2008, 12:50 PM
7 months until the competition guys.

If you've got a device capable of producing fire positioned under your rear, light it.

ahweh
01-14-2009, 02:25 PM
In the future will there be any Midwest events or regional preliminary events to make things more competitive on a local level.
Then the finals could be held in a different part of the country or world each year.

DresnerRobotics
01-14-2009, 02:46 PM
I am looking at the possibility of hosting a 2nd Mech Warfare event in the Midwest, near Chicago, however it would not likely be until 2010. I'm speaking with a few people in regards to bringing it to iHobby.

elios
01-14-2009, 05:01 PM
wow. thats cool

jes1510
01-14-2009, 10:28 PM
Now we're talking. That's driving distance!

ahweh
01-15-2009, 01:26 PM
Is there any possibility you might have a venue in the Kansas City area to consider because it is a major city as well.

DresnerRobotics
01-15-2009, 01:35 PM
In all honesty probably not. Two events a year requires a decent amount of resources and planning on my part, and we haven't even had the first year competition. If it were somehow to become insanely popular, then perhaps, but at this point I don't see it happening more than twice a year and those venues are already pretty much set. San Fransisco and Chicago are good locations for west, midwest, and east coasters.

In short, lets get year 1 launched and we'll play it by ear ;)

robologist
01-15-2009, 04:33 PM
Ya know, I always thought Dallas/Fort Worth to be a great mid-point for a gathering. :D

Dementedwombat
01-15-2009, 09:47 PM
Why don't we just host it in Lebanon Kansas (http://maps.google.com/maps?q=Lebanon,+Kansas&um=1&ie=UTF-8&split=0&sa=X&oi=geocode_result&resnum=1&ct=title) and make everyone drive the same distance :)?

elios
01-15-2009, 10:03 PM
thats good. "center of the US"

Adrenalynn
01-15-2009, 10:23 PM
You gonna front the money for the venue, the battlebox, insurance, etc? I'm wondering if in people's excitement they're forgetting that this is effectively costing them nothing or next to nothing outside their bots. It costs hundreds of thousands of dollars to host an event like this. Insurance alone is a gawdawful nightmare. Renting the venue is hardly easier. And $20k+ to put together even a small box.

Dementedwombat
01-15-2009, 10:34 PM
Not intended seriously, although I am a firm believer in the "open field and plywood" school of safety design.


You gonna front the money for the venue, the battlebox, insurance, etc? I'm wondering if in people's excitement they're forgetting that this is effectively costing them nothing or next to nothing outside their bots. It costs hundreds of thousands of dollars to host an event like this. Insurance alone is a gawdawful nightmare. Renting the venue is hardly easier. And $20k+ to put together even a small box.

Adrenalynn
01-15-2009, 11:10 PM
You and me both. Alas, I wouldn't get within a thousand miles of it these days. Once someone has something for someone to sue one for, one tends to avoid stuff like that...

I can almost guarantee you whoever hosts that will be out millions by the time their done... In the litigious world we live in, it just aint worth it.

Then after everyone is done suing you, you get thrown in prison for willful negligence. Just to add insult to injury

DresnerRobotics
01-15-2009, 11:48 PM
Yar good points. One of the reasons we're hosting it at Robogames and looking at the possibility of hosting it at iHobby is because its what I'm able to do given my resources. As much as I'd love to do this more than once or twice a year, my time and resources are limited so we've gotta work with what we have ;)

The target scoring system, the arena, the battlebox, and the A/V system to record our fights isn't exactly an easy thing to transport around.

lnxfergy
01-16-2009, 01:14 AM
You and me both. Alas, I wouldn't get within a thousand miles of it these days. Once someone has something for someone to sue one for, one tends to avoid stuff like that...

I can almost guarantee you whoever hosts that will be out millions by the time their done... In the litigious world we live in, it just aint worth it.

Then after everyone is done suing you, you get thrown in prison for willful negligence. Just to add insult to injury

Clearly, the event can only proceed once the event organizer has built a robot that is capable of killing anyone who thinks of bringing suit against them.... Tyb should be ready with that in what, like 3 months....

-Fergs

Adrenalynn
01-16-2009, 01:40 AM
I'm loaning him a dot50...

FryGuy
01-16-2009, 03:29 AM
You gonna front the money for the venue, the battlebox, insurance, etc? I'm wondering if in people's excitement they're forgetting that this is effectively costing them nothing or next to nothing outside their bots. It costs hundreds of thousands of dollars to host an event like this. Insurance alone is a gawdawful nightmare. Renting the venue is hardly easier. And $20k+ to put together even a small box.

I think robogames runs in the 20-40k range on a per-event fee (venue, security, insurance, maintenence, etc), not in the 200-400k range. I'm not Dave though, so I don't know for sure.

Hosting something smaller for just a mechwars event I think would be in the $500-5k range. Even cheaper if you can get the venue for free (such as at a hobby shop); sometimes their umbrella insurance will cover the event as well. I know that Sacramento State holds an annual antweight competition ("engineering expo"), and it only costs about $400 to rent the hall and about $200 for insurance.

DresnerRobotics
01-16-2009, 08:43 AM
Clearly, the event can only proceed once the event organizer has built a robot that is capable of killing anyone who thinks of bringing suit against them.... Tyb should be ready with that in what, like 3 months....

-Fergs

LOL. Truth!

Adrenalynn
01-16-2009, 03:00 PM
Flying projectiles with insurance for $200? Could you link that for me?

How much did the box build to cost? Dave posted it a couple times. As well as what he was willing to sell it for.

Could you show me how to ship that across country for less than a hundred dollars? (given your $500 total estimate and subtracting your insurance and venue)

FryGuy
01-16-2009, 05:50 PM
Flying projectiles with insurance for $200? Could you link that for me?

How much did the box build to cost? Dave posted it a couple times. As well as what he was willing to sell it for.

Could you show me how to ship that across country for less than a hundred dollars? (given your $500 total estimate and subtracting your insurance and venue)

I don't have any ability to link insurance, because it's all done on a quote basis. And my $200 was for an antweight event, which I feel is similar enough to this kind of event. Consider it's a single-day event, and and the projectiles are flying within a sealed environment, and it's relatively safe. At least, I would consider it equal in safety, or more safe than a 60 pound combat robot inside an arena. There is no chance of a 60 pound spinning weapon of death to breach the walls. And also, like I said, you could possibly get "free" insurance through the venue you're hosting it at, which would put the cost at the low end of the range.

The Sacramento State's antweight arena (6'x6'x4') cost about $2500 to build, 4-5 years ago. Given that the cost of an arena could be amortized over several events, it's not really that much, compared to the $1000+ it takes each competitor to make their mech.

As for transport, I'm assuming that the event would be local to whoever was holding it. Meaning, they could stick it in their truck and drive it to the venue. Obviously, having to pay a shipper wouldn't put the price at the absolute lowest of the range I suggested.

Heck, one of my favorite antweight events was called "Marin Ant Wars: Robots & Ribs," and was held in a builder's garage in Marin County. He invited builders and their friends to come and compete, and in the break he served a barbeque lunch. I'm just saying you don't have to be hosting Robogames to host an event, and that your ballpark minimum of 100k+ is off by a factor of 100.

ahweh
01-17-2009, 11:02 PM
Lebanon Kansas is only about 2 hours from where we live here in Atchison which is why I asked about Kansas City.
So do you live in Lebanon Kansas or do you just like the place.
George

Dementedwombat
01-18-2009, 11:51 AM
I just picked it because it is the approximate center of the U.S. That way no one can complain about it being out of their way.

Adrenalynn
01-18-2009, 11:54 AM
No one SHOULD complain either way. That's like whining about the gift someone gives you, ya know?

Dementedwombat
01-18-2009, 12:00 PM
I agree, but I've discovered that it's human nature to complain. Just trying to reduce the complaints as much as possible (then again setting the event in Lebanon might start a whole new set of complaints :)).

Adrenalynn
01-18-2009, 01:23 PM
Maybe there's an easier way. "Anyone that doesn't want to come to the event being held where it's being held may, with properly given notice, elect not to"?

FryGuy
01-19-2009, 01:50 PM
Why do they need to give notice of not coming? Seems backwards :p

Adrenalynn
01-19-2009, 02:37 PM
Well, we need to know how may people aren't going to show up, of course! :tongue:

DresnerRobotics
11-17-2009, 09:40 AM
Heads up guys, now less than 6 months until Robogames in April. Get your butts into gear if you plan on competing in 2010.

xdream
11-24-2009, 01:12 PM
X-Brat is coming along nicely! Hope to make more progress over the long Thanksgiving weekend.
Mark