Rawbots Club Forum

Rawbots General => Challenges => Topic started by: PressureLine on March 16, 2016, 10:13:13 am

Title: Autonomous Combat League
Post by: PressureLine on March 16, 2016, 10:13:13 am
Welcome to the Autonomous Combat League!

This thread is dedicated to battles between autonomous bots in Rawbots. It is intended to be a fun and friendly competition between bots and bot builders while providing a bit of spectacle to the Rawbots Club community. Building and programming an autonomous bot requires a fairly high level of effort, so be respectful to your fellow Rawboticists and above all have fun!
 
Competition/Construction Rules:

Classes:
To counter the trend towards increasing bot sizes and combat power it has been decided to class ACL bots by partcount:

Suggestions:

Creating Arenas:

Running battles:

It is probably worth keeping this thread for entry announcements and results discussions. Please use the Combat Bot Workshop (http://forum.rawbots.club/index.php?topic=37) thread for discussing the construction and programming of competing bots (feel free to debate the merits [or otherwise] of bots and arenas here though).

*edit* 28/03/16 - Change to weight class rules
Title: Re: Autonomous Combat League
Post by: PressureLine on March 16, 2016, 10:15:30 am
Unlimited and Legacy Bots:

A listing of combat bots built with over 75 parts, and bots that otherwise do not fit into the construction rules as laid out above. Please be aware that the 'Legacy' bots may not work all that well in Rawbots + XFM 0.5 due to changes in flux capacitor recharge times, and other changes to cameras and hooks.

Unlimited: >75 parts

Legacy
Title: Re: Autonomous Combat League
Post by: PressureLine on March 16, 2016, 10:19:34 am
Competition Entries:

A listing of combat bots by class:

Lightweight: <=25 parts
Middleweight: 26-50 parts

Heavyweight: 51-75 parts
Title: Re: Autonomous Combat League
Post by: PressureLine on March 16, 2016, 10:25:16 am
Competition Arenas:

Arenas built for ACL competitions by author:

PressureLine (http://forum.rawbots.club/index.php?action=profile;u=1):

MarvinMan (http://forum.rawbots.club/index.php?action=profile;u=8):

tob.s:
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 16, 2016, 02:19:13 pm
It's great to have the ACL officially running again. I'll claim blue again as my team colour.

Here's my old combat bot, probably still valid for light or middle weight. Does have non-standard enable controls (g and h for drive and weapons). It was the original proof of concept bot, and by the end of the last ACL round it was entirely outclassed by everything else around. Should still make a lightweight decent target for tuning new bot designs.
https://www.dropbox.com/s/lgkm5u73wdo35kl/bp_autoX3?dl=0 (https://www.dropbox.com/s/lgkm5u73wdo35kl/bp_autoX3?dl=0)

This is my original arena, with 6 starting garages and a central platform.
https://www.dropbox.com/s/hkxm8qur4sh1o5g/arena.zip?dl=0 (https://www.dropbox.com/s/hkxm8qur4sh1o5g/arena.zip?dl=0)

This is a new arena "The triangle", designed for 3 players with a raised region in the centre.
https://www.dropbox.com/s/dmv8l4pjlsqd2fj/acl.zip?dl=0 (https://www.dropbox.com/s/dmv8l4pjlsqd2fj/acl.zip?dl=0)
Title: Re: Autonomous Combat League
Post by: z26 on March 16, 2016, 07:29:30 pm
Hi, so this also is the discussion thead about anything acl related, not only important stuff like bot registration right?

also, "Some arenas may not have detectable floors, please make cliff/edge detection systems on/off-able."

Does that still apply now that energy bridges are attachable by hooks?  I guess there are niche cases like small floor holes or continuum platforms, but these aren't frequent.
Title: Re: Autonomous Combat League
Post by: PressureLine on March 16, 2016, 08:48:08 pm
Hi, so this also is the discussion thead about anything acl related, not only important stuff like bot registration right?

Yep, but I will start up a workshop topic to talk about bot building and coding.

also:

Code: [Select]
"Some arenas may not have detectable floors, please make cliff/edge detection systems on/off-able."
Does that still apply now that energy bridges are attachable by hooks?  I guess there are niche cases like small floor holes or continuum platforms, but these aren't frequent.

Not really applicable now, but in all that recycled text I was bound to miss something right?

edit: also put MarvinMan's bot and arenas in the directory posts :)
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 16, 2016, 09:15:50 pm
I've been working on a new navigation system as my old bot (and several others) was rather slow. This one won't get stuck in tight corners, and can drive up and down ramps. It can't detect if its about to drive off a cliff though, to the original arena could cause problems without a self-righting system.

Hopefully I should be able to add a weapons system soon to turn it into a fully functional combat bot.
(http://i.imgur.com/AMNrzSS.png)
Title: Re: Autonomous Combat League
Post by: z26 on March 16, 2016, 09:38:38 pm
so it can both assess the horizontal and vertical slope of whatever surface in front of him? clever.
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 16, 2016, 09:51:56 pm
The lower hooks are used for navigation as normal, and the bot tries to keep a distance of about 25 between the hook and the wall.

If there's a difference of more than ~1.5 between the top and bottom measurements, it can see a ramp and it ignores the distance readings from the bottom hooks.

The elbows oscillate to scan an area in front of the bot, so the upper hooks need to move with them.
Title: Re: Autonomous Combat League
Post by: PressureLine on March 18, 2016, 03:45:23 am
I've come up with a design that will be an entrant in the 'Lightweight' class, which I will also try to convert into a 2-bot middleweight entry.
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 18, 2016, 08:38:03 am
I'm aiming for the lightweight class too with my current design. The prototype will use plasma cannons, but I'm working on shrinking my grenade cannon down to a usable size.

Using plasma cannons as a propulsion source for grenades seemed like a good idea, but it also activates their cooldown timer.

Here's my grenade cannon with a 6 round autoloader. It's ripped straight out of my amphibious tank and had to be supported with hooks, so it will spawn one hex above the floor of a blueshift map. Press A to fire, and hold R until the blue light comes on to reload.
https://www.dropbox.com/s/ztc3ojvuldg6byo/bp_g_gun?dl=0 (https://www.dropbox.com/s/ztc3ojvuldg6byo/bp_g_gun?dl=0)
Title: Re: Autonomous Combat League
Post by: PressureLine on March 18, 2016, 11:10:58 am
Figured it was about time I had a play around with the big terrain hexes :)

(http://i.imgur.com/7dbVZTi.jpg)
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 18, 2016, 11:24:10 am
That looks like a pretty evil test of a bots navigation abilities.

My combat bot is slowly getting there. So far I've had it pursuing a pressuretank around the map, but it still needs a bit of work to make the target following and navigation systems interact properly.
Title: Re: Autonomous Combat League
Post by: PressureLine on March 18, 2016, 11:25:28 am
(http://i.imgur.com/UMzT9vm.jpg)
(http://i.imgur.com/9L4oLDp.jpg)

It is a brutal close combat arena, make sure your bot can flip back onto it's wheels (or like Tumbler is invertable)
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 18, 2016, 11:30:46 am
Wow, it's not often that you see a defeated combat bot (or two?) in that many pieces.

A self righting or stabiliser system is definitely on the to-do list, as my bot manages to roll without going at full speed.

I was going to put a couple of high power cannons on the front of the bot, but seeing that, lasers might be a better option.
Title: Re: Autonomous Combat League
Post by: PressureLine on March 18, 2016, 11:41:53 am
Lasers tend to do that, especially when the drive system of the bot keeps driving the attacker into (and often through) it's target.

FireStorm (which is a similar design to Tumbler, but not invertible) is fairly effective too, with 3x 1 capacitance cannons fixed forwards and essentially the same drive code it'll just ram it's targets at full speed while pouring plasma into it (which is why I listed it as 'Plasma Melee') Which while not as effective at total dismemberment as a laser bot is, still has it's merits as an attack mode (since it can still hit it's target at range to cause enough damage to make it easier to finish off it's opponent at melee range)
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 18, 2016, 11:51:16 am
If fixed lasers are that effective, I wonder what spinning lasers could do.

I suppose those ramming tactics become less effective when the opponent is trying the same thing.

I'm planning a slightly more sophisticated strategy of charging at the opponent, dumping several high energy rounds into it then retreating to recharge/reload.
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 19, 2016, 12:37:50 am
Right, my combat bot now has a functional weapons system. It's got no armour yet, so it will probably be destroyed by anything with ranged weapons. I put it up against Pressurelines Sawtooth, and that made for a rather interesting match.

Here's the blueprint. I've named it the broadside, on account of the fixed side-facing cannons and its strategy of dumping everything its got into the target as it comes along side.
https://www.dropbox.com/s/n4p2819hvxu65ca/bp_broadside?dl=0 (https://www.dropbox.com/s/n4p2819hvxu65ca/bp_broadside?dl=0)

Match summary:
>First blood goes to the broadside, knocking off a couple of floaters.
>Gets a few shots in while staying out of the Sawtooths attack range.
>The battle starts to get into a tight corner, hindering the broadsides ability to keep away.
>Things get rather messy, with both bots rammed in a corner.
>Eventually ends up with a head-on collision, resulting in the broadside loosing a cannon and a hook.
>Overall outcome is the Sawtooth stuck on its side against a wall, and the Broadside driving around completely ignoring the stationary target.
(http://i.imgur.com/eoCLqgw.jpg)
Title: Re: Autonomous Combat League
Post by: PressureLine on March 19, 2016, 02:04:09 am
Firestorm vs SawTooth vs Broadside

(http://i.imgur.com/MGHYItu.jpg)
(http://i.imgur.com/2QxYW9s.jpg)
(http://i.imgur.com/x77lk4h.jpg)

Might have been the arena (lavafield) but Broadside never quite seemed to really get into it. Especially after a close encounter with SawTooth's lasers which detached one of it's wheels and caused it to spend a lot of the rest of the match using the jet to right itself until FireStorm caught it in one of the start 'garages' and finished the job.
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 19, 2016, 09:00:57 am
Actually, I've been seeing the same sort of thing on all the matches I've tried after that. I know it doesn't handle multiple opponents well, especially if they're not moving fast enough to be detected.

Most of the problem seems to be that it's getting too close to the target and not circling around it. Hopefully that will be resolved somewhat if it reverses the drive when it detects it's too close to both the terrain and  its target.

Perhaps I'll just go with a strategy that doesn't require avoiding collisions with the target.
Title: Autonomous Combat League - Goblin
Post by: PressureLine on March 22, 2016, 09:43:42 am
New Lightweight competitor!

Goblin

(http://i.imgur.com/mclLr4c.jpg)

Weighing in at just 23 parts, Goblin is a lightweight Plasma Shooter capable of engaging both ground based and aerial opponents.

Get Goblin here (https://drive.google.com/open?id=0B42Wtrlp8xi9akdEQTNPYjZVbUk).
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 22, 2016, 12:57:58 pm
That thing's pretty cool and works well, especially considering how few parts it uses. I had a couple of test matches, and one of them even managed to damage a PressureTank. The only issue I've noticed is that they like getting stuck either upside-down or on loose continuums. The cannon looks as though it would work as a self-righting mechanism without adding any parts.

UPDATE: I've just seen the Goblin score an absurdly lucky shot against the SawTooth, putting a shot right between the floaters on the front armour to hit the elbow that the lasers are mounted on. It managed to do that right as the lasers were being deployed, resulting in them being thrown away.
Title: Re: Autonomous Combat League
Post by: PressureLine on March 28, 2016, 05:21:58 am
Ok, minor rules change to spread the weight classes around a bit:
There is a bit of reasoning behind this, so bear with me: It makes each weightclass a bit broader, and better represents (imo) the way combat power escalates with increased partcount. It should also have the effect of making any potential swarmbots a bit more competiticve, since now a combo of 2 max weight light bots counts as a single middleweight entry, and not a heavyweight.
Title: Re: Autonomous Combat League - autoX3_acl
Post by: PressureLine on March 28, 2016, 06:40:45 am
'New' middleweight competitor!

autoX3_acl

(http://i.imgur.com/M21fVKO.jpg)

A minor rebuild of MarvinMan's (http://forum.rawbots.club/index.php?action=profile;u=8) original combat bot, autoX3_acl is a middleweight Plasma Shooter capable of engaging both ground based and aerial opponents.

Get autoX3_acl here (https://drive.google.com/open?id=0B42Wtrlp8xi9Mnc2NWpxS0E2Y00).
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 28, 2016, 11:11:57 am
That looks good. I wonder how well it'll hold up only being able to fire backwards while being one of the slowest bots out there?

In other news, the chassis for my new combat bot is nearly done, then I'll just need to transplant the code over from the test platform and it'll be ready to compete.
Title: Re: Autonomous Combat League - Canary
Post by: PressureLine on March 28, 2016, 11:18:22 am
New middleweight competitor!

Canary

(http://i.imgur.com/2NP8xdu.jpg)

Canary is a middleweight hovering Plasma Shooter capable of tracking engaging ground based targets but has limited function against other aerial targets. The black/yellow block is the altitude control block, when detached it becomes extremely heavy and almost impossible to move, detaching this block also triggers the lift system to engage.

Get Canary here (https://drive.google.com/open?id=0B42Wtrlp8xi9Rl9sZFk2Q1l2b3c).
Title: Re: Autonomous Combat League
Post by: PressureLine on March 28, 2016, 11:36:45 am
That looks good. I wonder how well it'll hold up only being able to fire backwards while being one of the slowest bots out there?

In this case:

(http://i.imgur.com/0pE8aOP.jpg)

Not well. That being said, while doing the run-in testing for Canary I've seen shots sail through the armor and jets and squarely strike the continuum at the back of the upper turret causing the whole turret and aiming system to fail, so sometimes it just comes down to luck.
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 28, 2016, 11:44:50 am
Well some of the parts are still connected ;D

The turret on my old bot does do the full ballistic equations and leads the target, so it should still be fairly competitive if it survives long enough to acquire a target. However, I assume it's no longer calibrated right after all the tweaks to plasma cannons, and IIRC it still uses an integrator to do the inverse trig functions, so it probably doesn't handle fast moving targets well.
Title: Re: Autonomous Combat League
Post by: PressureLine on March 28, 2016, 11:58:30 am
If you are using shoot_force as the primary method of working out the speed of the plasma bolts you should be about fine, that relationship hasn't really changed, the only difference is how much force you get per capacitance.

Also! 'New' arena up: Tech Arena Hex (http://forum.rawbots.club/index.php?topic=50). It's a big, open arena, will definitely suit fast bots with decent target following abilities more than the bots we were running back in 2014 where on an arena this large you could end up waiting for a long time before the bots would actually engage.
Title: Re: Autonomous Combat League
Post by: PressureLine on March 28, 2016, 01:17:38 pm
Just found this in my blueprints:

(http://i.imgur.com/cjXBJl3.jpg)
(http://i.imgur.com/AvEsKJP.jpg)
(http://i.imgur.com/eTdsTvQ.jpg)

Made by tob.s it's a second version of his 'erazer' bot. Incredibly fast, armed with three lasers on the front and a ramming attack that has to be seen to be believed. No idea what the partcount is, but it's gotta be around 100.
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 28, 2016, 08:21:27 pm
I don't remember that one. Could you put the blueprint up? If its charging attack is as bad as you say, it should make for a decent opponent to test a collision avoidance system against.
Title: Re: Autonomous Combat League
Post by: PressureLine on March 28, 2016, 10:57:12 pm
Well it has a ground speed of 40-50, so it's ramming/charging attack usually results in the target being thrown a good distance (if the lasers don't cut it in half) when it hits. I'm not sure how complete the code is, since it doesn't seem to track the target very well, mostly concentrating on hooning around the arena looking awesome.
Title: Re: Autonomous Combat League
Post by: MarvinMan on March 30, 2016, 03:17:02 pm
Here's a sample of what I've been working on- the RedEye. The mechanical structure and navigation systems are done, and the weapons system makes an attempt at hitting its target. In my testing, it managed to do a reasonable amount of damage to the PressureTank, but was usually disarmed or disabled before immobilizing it.

The turret still needs work as it only covers a 180 degree range (it least it's the front 180 degrees this time), doesn't make any attempt to lead its target and has no elevation control. It would probably be a good idea to add a collision avoidance system too, so it doesn't charge head first into the first laser bot it comes across.

I haven't been able to get a stable or accurate part count out of a camera, but I counted it at 51, which puts it in the unfortunate position of being the smallest and least armored bot in the heavyweight class.

Blueprint: https://www.dropbox.com/s/h4nivwg9rfypqdz/bp_redeye?dl=0 (https://www.dropbox.com/s/h4nivwg9rfypqdz/bp_redeye?dl=0)
(http://i.imgur.com/EXeQwI1.jpg)
Title: Re: Autonomous Combat League
Post by: PressureLine on March 30, 2016, 11:17:02 pm
Looks good, can't wait to try it out. I will also update the directory post once I get home :)
Title: Re: Autonomous Combat League
Post by: PressureLine on March 31, 2016, 12:58:12 am
I haven't been able to get a stable or accurate part count out of a camera, but I counted it at 51, which puts it in the unfortunate position of being the smallest and least armored bot in the heavyweight class.

FWIW I counted it at 51 too. It can join your original autoX3 (52) and Sawtooth (51) in the light-heavyweight category.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 01, 2016, 01:51:05 am
Almost finished a mediumweight ground bot. I just want to get a workable 'approach but don't ram' system working :)
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 01, 2016, 11:12:18 am
As much as the target following behavior makes bots engage more quickly and keep attacking, the ramming behavior does somewhat ruin it as the winner will usually be the bot that ends up on the bottom (or has lasers). At least with random walk behavior, you tended to get multiple encounters between multiple bots with relatively few collisions, even if it took a lang time to get a complete kill.

I'm planning to add a system to keep an optimal distance to the target, based on what the broadside had,  but I can see the situation getting messy for bots designed to go forward only, especially in tighter parts of the arena. I'm thinking that adding a second set of rear-facing hooks to do collision avoidance might let the bot get away from its enemy more easily. Kinda makes me wish I'd kept the whole reversible drive and navigation system.
Title: Re: Autonomous Combat League - HailStorm
Post by: PressureLine on April 01, 2016, 01:25:21 pm
New middleweight competitor!

HailStorm

(http://i.imgur.com/yVctHq1.jpg)

Weighing in at 44 parts, HailStorm is a middleweight Plasma Shooter capable of tracking and engaging both ground based and aerial opponents.

Get HailStorm here (https://drive.google.com/open?id=0B42Wtrlp8xi9R2M1TzItMm43Mm8).
Title: Re: Autonomous Combat League
Post by: PressureLine on April 01, 2016, 01:27:26 pm
As much as the target following behavior makes bots engage more quickly and keep attacking, the ramming behavior does somewhat ruin it as the winner will usually be the bot that ends up on the bottom (or has lasers). At least with random walk behavior, you tended to get multiple encounters between multiple bots with relatively few collisions, even if it took a lang time to get a complete kill.

I'm planning to add a system to keep an optimal distance to the target, based on what the broadside had,  but I can see the situation getting messy for bots designed to go forward only, especially in tighter parts of the arena. I'm thinking that adding a second set of rear-facing hooks to do collision avoidance might let the bot get away from its enemy more easily. Kinda makes me wish I'd kept the whole reversible drive and navigation system.

The system in HailStorm won't work for all bots, but works pretty well with it's turret design, I'm happy with it anyway.
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 01, 2016, 03:07:17 pm
The HailStorm is a very close match for my RedEye, and the winner seems to depend on how they encounter each other. If the HailStorm gets line of sight from a distance or there is a significant altitude difference, it usually has the advantage. If the bots meet at closer range, the RedEyes very basic aiming is much less of a disadvantage and the Hailstorm tends to drive over it or roll, exposing its vulnerable underside.

One other thing I've noticed is that the turret assembly appears to be fairly easy to dislodge, although that could be because the RedEye can only really hit things high enough to line up with its own turret.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 02, 2016, 04:14:02 am
Did a quick 3-way between Erazer, HailStorm and RedEye:

Spoiler: show

(http://i.imgur.com/t6YcCua.jpg)
(http://i.imgur.com/UOyfci1.jpg)
(http://i.imgur.com/0cCYDE1.jpg)
(http://i.imgur.com/bYL2G83.jpg)
(http://i.imgur.com/zq0pWkq.jpg)
(http://i.imgur.com/y3WQZC2.jpg)
(http://i.imgur.com/CHJm9V3.jpg)
(http://i.imgur.com/qa1wKMv.jpg)


And the final results:
RedEye: show
(http://i.imgur.com/s2RCVqo.jpg)

Erazer: show
(http://i.imgur.com/a8cGoN5.jpg)


And the winner:
HailStorm: show

(http://i.imgur.com/xeSUFCa.jpg)
(http://i.imgur.com/wInPFzH.jpg)


I think RedEye got caught by Erazer's lasers early on which cost RedEye it's two front wheels. Hailstorm added a bit more damage, before Erazer returned to finish the job. after that was a bit of back-and forth cannon fire between Erazer and HailStorm until all of Erazer's weapons were destroyed.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 02, 2016, 05:54:46 am
So I added cannon elevation control to RedEye (was easy once I figured out the 'part' output flow) and this is the result:

(http://i.imgur.com/bu6btGZ.jpg)

(Yeah, thats HailStorm in tiny pieces, a turretless Canary and an undamaged RedEye)
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 02, 2016, 11:33:33 am
Wow, I wasn't expecting a properly set up turret to make that much difference. Without it, I was finding it being 50-50 whether the RedEye or Hailstorm would win, and that was in a relatively tight arena where the Hailstorm seemed somewhat prone to getting stuck or rolling.

Regarding the first round with the erazer, I had a similar result against the SawTooth. The RedEye normally seems to take losing an axle quite well, it's the loss of the motors and camera directly behind it that disable it as they deal with moving the turret and deciding whether to enable the cannons.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 02, 2016, 12:27:58 pm
To be fair, Canary and HailStorm were already engaging each other when RedEye joined the party, and it's hard to keep track of which bot actually did the most damage when it's a fairly short-ranged gunfight, but RedEye definately fired the killing blow on at least one of the two (although I'm pretty sure it was on both)

I have been going through my old blueprints, and it's actually really cool to see how (at least for my bots) the code has evolved over time as I have built them.

Still trying to think of the next 'gimmick' for my next bot, might try and give grenades a go.
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 02, 2016, 08:50:37 pm
Yeah, it is quite amusing how my first combat bot doesn't stand a chance against any of the later bots.

I had been considering building a bot with a number of single-use lances tipped with a grenade. The lances would either be made from pistons or elbows held together with long arcs so they can be stored in a compact form and then hold the grenade away from the bot when deployed. Depending on the target part count and bot size, the lances could be kept in an armored box.

Firing grenades with a plasma cannon looked like a nice idea, but it triggers the cooldown on the grenade.
Title: Re: Autonomous Combat League
Post by: z26 on April 02, 2016, 09:28:46 pm
As Pressureline told me thats the unfortunate consequence of energy and hp being the same thing.  Low hp = depleted energy.
Title: Re: Autonomous Combat League
Post by: Sleeping_Owl on April 03, 2016, 07:06:51 pm
One of my designs from old forums
Nest (unlimited)
I didn't count parts, but this one is so fat it could be the reason why bots got divided into classes.
Iron-shaped war machine with three land missiles. If it can't destroy your bot, it will destroy your CPU.

(http://i.imgur.com/yX1Q7xu.png)

https://drive.google.com/open?id=0ByzpYkai1-8lYkhLLWkxdVlRQkU (https://drive.google.com/open?id=0ByzpYkai1-8lYkhLLWkxdVlRQkU)

How to count parts? Camera's activity underestimates this thing.
Title: Re: Autonomous Combat League
Post by: z26 on April 03, 2016, 07:21:39 pm
How to count parts? Camera's activity underestimates this thing.

Carefully :p

Oh and welcome!
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 03, 2016, 11:04:02 pm
Is there any way to beat that with the current rules without resorting to grenade based countermeasures? Ramming attacks seem to be very effective despite the nerfed lasers, so two grenades should take out practically anything. There aren't any bots fast enough to evade the missiles, and deploying decoys isn't really practical, so the only real option left is to try to intercept the missiles using cannon fire, requiring high DPS and/or high accuracy.

Seeing how well the current ACL format is working, I was thinking about starting up an aircraft version. In theory, the lack of terrain makes it easier to get into, but the full 3D flight makes it much harder to perfect.
Title: Re: Autonomous Combat League
Post by: z26 on April 03, 2016, 11:10:40 pm
a purposefully made hovering bot might be able to evade these missiles, since it doesn't have to worry about obstacles as much when moving around.
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 03, 2016, 11:27:47 pm
Depending on how the missiles are triggered, a hover bot might not have to deal with them at all. Dodging the initial ramming probably isn't too difficult, but persistently keeping out of the blast radius as the missile follows the bot around would be near impossible. And that's before you consider that there may be multiple missiles or other opponents.
Title: Re: Autonomous Combat League
Post by: Sleeping_Owl on April 04, 2016, 12:26:00 am
Nest can attack only ground targets, so there is an obvious way to defeat it - fly (such fight would be boring).

There still is no missile that can effectively destroy both air and ground targets.
My experiments stopped at three branches:
1. Land missile.
2. AA missile that can't hit ground targets.
3. Flying missile that barely hits ground targets, useless against moving targets.

So missiles are only good in unlimited class, they are specialized and you can't have more than 2-3 of them.

I think it is possible to build plasma bot with better aim that can dodge for some time and damage any existing missile enough to make it useless. We haven't seen fully armored missiles yet.
Such bot would require quite much math, and rawbots has terrible interface for math.
If it will be a ground bot, it should not use wheels, jet slider design is better at dodging because it can move in any direction.

So given enough effort, Nest can be beaten with plasma. But if an arms race starts, I would bet on missile bots in 1 vs 1, and grenade-ramming swarms in free fights.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 04, 2016, 06:09:48 am
I didn't count parts, but this one is so fat it could be the reason why bots got divided into classes.

...

How to count parts? Camera's activity underestimates this thing.

I load up the blueprint in Notepad++ and get it to count all the instances of "local_to a_" then subtract any control blocks, as well as all the instances of "_rotor\n" (for the second part of motors and elbows) and "_head\n" (for the second part of pistons) remember to change the search type to the 'Extended' type (\n is for new line)
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 04, 2016, 10:13:14 am
Perhaps anti-air missiles would be more effective if they got a burst of acceleration when they were exactly aligned with the target. That should help with attacking ground based targets too, as the missile could align itself at a relatively high altitude, then perform a high-speed dive into the target.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 04, 2016, 10:16:59 am
I get the raw partcount of Nest as 189, once the control block and the second parts of any motors, elbows and pistons are subtracted it comes out to 170. I'll update the bot directory post accordingly.
Title: Re: Autonomous Combat League
Post by: cupid_the_conqueror on April 04, 2016, 05:31:36 pm
Perhaps anti-air missiles would be more effective if they got a burst of acceleration when they were exactly aligned with the target. That should help with attacking ground based targets too, as the missile could align itself at a relatively high altitude, then perform a high-speed dive into the target.

this is really tricky though, I tried to make a short range rocket that would just accelerate super fast to hit its target, and it would either be tooo slow, or clip through the target, or just knock the target flying with the kinetic force before the grenade exploded.

if your traveling at more then 200 velocity, you'll often clip right through the hexterrian
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 04, 2016, 08:58:21 pm
it would either be tooo slow, or clip through the target, or just knock the target flying with the kinetic force before the grenade exploded.
Perhaps kinetic weapons could be effective in some cases. Many of the combat bots are as good as dead once they roll, so a kinetic missile could be used to flip a bot before finishing it off with conventional weapons. They may work as a countermeasure against ramming attacks too.
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 04, 2016, 11:44:17 pm
I've been playing around with swarm bots a bit more. I re-built my previous swarm bot with hooks for navigation and copied the control code across from my main combat bot. The result makes much better progress around the map than the previous version that bounced off of the walls.

I'm not sure what to do about weapons as there's only one free connection point left. I was going to try putting a grenade on each bot, but lasers could work well and would take less effort to coordinate the attack. I suppose I could try and put a cannon on it again, but that causes a load of problems with aim, recoil and where to put the capacitor.

BTW, does projectile velocity affect damage? I was playing around with boosting the velocity of low energy shots, and they seemed to be doing more damage than expected.
(http://i.imgur.com/fsQw5gb.jpg)
Title: Re: Autonomous Combat League
Post by: Sleeping_Owl on April 05, 2016, 12:00:48 am
I load up the blueprint in Notepad++ and get it to count all the instances of "local_to a_" then subtract any control blocks, as well as all the instances of "_rotor\n" (for the second part of motors and elbows) and "_head\n" (for the second part of pistons) remember to change the search type to the 'Extended' type (\n is for new line)

Here is a script that does what you described. Needs either linux shell or cygwin. Says that Nest is 171 part.
Use like:
$ ./count_parts.sh < bp_Nest

https://drive.google.com/open?id=0ByzpYkai1-8ldE1JZ0xmQWNtMFE (https://drive.google.com/open?id=0ByzpYkai1-8ldE1JZ0xmQWNtMFE)
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 05, 2016, 12:44:04 pm
These laser-armed swarm bots (I call them ants) are very effective against slow, armored targets. I put 3 of them (21 parts total) running completely independent code in the arena with a PressureTank, and one of them was the last bot left intact. Two of the ants got in a bit of a friendly fire incident and completely disassembled each other, but still managed to distract the PressureTanks turret, allowing the third one to get close enough to remove a wheel and do some serious chassis damage.

The surviving ant did beach itself on the top of a ramp after that, but was otherwise intact and functional.
(http://i.imgur.com/NLyGsS7.jpg)

After being unstuck, the ant went back to achieve mutual annihilation of itself and the PressureTank.
(http://i.imgur.com/i95ToKR.jpg)

The next thing to do is tweak the mechanical design so it can deal with ramps and add a bit of coordination between multiple bots. I might have a match against the erazer too, as I suspect the result will be interesting.

UPDATE: 3x ants V erazer resulted in a decisive victory for the ants, with 2 fully functional survivors.
(http://i.imgur.com/LsGESlD.jpg)

Does anyone know why lasers don't keep the colour they were set to? The lasers on the ants start out blue, but change to the default red after a while.
Title: Re: Autonomous Combat League
Post by: z26 on April 05, 2016, 01:50:00 pm
idk, but in my case when i wanted a "black" laser (actually a dimmer white) it wouldn't keep the color so i used a cyclically activated multiplexer to refresh it, so that one potential way to deal with the issue.
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 05, 2016, 02:50:23 pm
I've already got a rapidly cycling toggle for the camera system, so it should be easy to add colour I/Os for the laser.

Here's the blueprint for the laser ant (V6), with slightly improved target acquisition system. The detection threshold is set to 8 parts to prevent targeting other ants, so this version may have problems against smaller bots. https://www.dropbox.com/s/nsedzsmt2zwwy7p/bp_ant6?dl=0 (https://www.dropbox.com/s/nsedzsmt2zwwy7p/bp_ant6?dl=0)

Does anyone have the blueprints for the later versions of the erazer? I'd be interested to see whether it has any better luck against the ants than the original. It seems as though while the erazer needs a large open map to work well, that gives all the ants a clear line of sight for them to all lock on at the same time.
Title: Re: Autonomous Combat League
Post by: z26 on April 05, 2016, 03:13:58 pm
so it seems like swarms bots are as effective as hoped, their main problem tho is being able to handle tricky terrain without being made of too many parts.

edit:  If swarm bots gets too annoying for big bots to deal with  (especially since given their low part count distinguishing them from debris is difficult) , maybe dividing the acl in two divisions, single bot and multi bots would be preferable.  Bot being defined as something that can propel itself in a controlled direction.  That would mean guided rockets would be considered bots.  Darts steered by fins might be tricky however.

If you could get the current hitpoints of a part (with the connectivity meter or something else) distinguishing between debris and targets wouldn't be as difficult.  Worst case scenario: if you target a healthy debris, instead of obsessing over it you only deplete its health to 0, then move on to another target.

Or maybe see when there is only a single input to the connectivity meter whenever that part is attached to anything at all whatsoever.  That could fulfill the same need.
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 05, 2016, 04:30:56 pm
The current ant swarm bot uses the almost same navigation system as my full combat bot, so they shouldn't have a problem with tricky terrain from a software perspective. The hooks need tilting upwards 20-30 degrees (like they are on the RedEye) to let it drive up ramps, and the center continuum needs shifting slightly to alter the balance and raise the ground clearance. The first ant with the cannon was perfectly capable of dealing with ramps, but had fairly serious aiming recoil problems.

Swarm bots seem to be less effective in the tighter, more complex maps where they are more prone to getting stuck on each other and it's easier to pick them off one by one. I like to filter detected parts to only lock onto parts moving faster than a threshold velocity, which should prevent the bot locking onto piles of debris while still being able to see small bots.

The PressureTank did seem to be getting close to hitting the ants, so some of the newer bots with better turrets might be able to fend them off reasonably well. Hover bots should be relatively immune being mobbed by swarm bots, given how much adding a cannon with elevation control would add to the part count and how much recoil affects such light bots. I don't know how well swarm V swarm matches would work as you'd need some sort of IFF system (the current ant bots work entirely independently and simply ignore anything ant sized or smaller), and any coordinated attacks would likely end up with every bot trying to pile into the same space.
Title: Re: Autonomous Combat League
Post by: z26 on April 05, 2016, 04:40:48 pm
I like to filter detected parts to only lock onto parts moving faster than a threshold velocity, which should prevent the bot locking onto piles of debris while still being able to see small bots.

Oh, that's good!  I'll use that.  Also, aiming indeed is an issue, but to handle recoil just increase the weight of a continuum for a couple frames when you're shooting.
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 05, 2016, 06:36:57 pm
I'll have to try increasing the mass when firing (although it is quite funny seeing the bot flying backwards several body lengths or even getting some air time every time it fires a high power round).

Velocity filtering can be a bit unreliable as it takes quite a while for the calculations to complete relative to how quickly the rest of the latch works. Often, the velocity calculation will keep the latch enabled after the camera has lost sight of the part, resulting in no target part being stored in memory and the bot wandering aimlessly. Adding an extra latch directly connected to the cameras detected and activity outputs seems to fix this.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 06, 2016, 05:00:26 am
To be fair, the turret on PressureTank turns far too slow afair, which causes it's accuracy to be poor because it is actually fairly fast (fast enough to generate misses against a stationary target)

I've got plans for another plasma-turret middleweight bot  that looks to be an interesting build at least :)
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 06, 2016, 09:44:43 am
The turret doesn't seem to be too slow now that bots tend to follow their target, as the target bot will usually approach or follow head on. A lot of turrets seem to have difficulty hitting low targets and end up firing over the top of them (which would be center-mass hits on a more sensible sized bot), especially at close range.

Good luck with the new bot, I'm still waiting for someone to build one with multiple independent turrets.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 06, 2016, 10:17:41 am
The turret doesn't seem to be too slow now that bots tend to follow their target, as the target bot will usually approach or follow head on. A lot of turrets seem to have difficulty hitting low targets and end up firing over the top of them (which would be center-mass hits on a more sensible sized bot), especially at close range.

Thats what the shape of HailStorm was all about. I think I managed to be able to allow -20° on the cannons of Hailstorm, which allows it to shoot very close to itself without shooting itself.

Good luck with the new bot, I'm still waiting for someone to build one with multiple independent turrets.

Considering that the code will be a direct port of the HailStom code, it's more an attempt to see how different construction philosophies affect the effectiveness of the bot. I did start on one flying bot with independent turrets, but it's too big and probably wouldn't be that effective.
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 06, 2016, 12:54:20 pm
I ran some big matches in an open arena, with the PressureTank, RedEye, SawTooth, HailStorm, erazer and some ants. The erazer usually went first, getting caught in the cross fire and apparently being the ants favorite target. The SawTooth usually did the mast damage, almost single-handedly winning one of the rounds. The PressureTank generally stayed out of the close range mess that the newer bots got into at the start, and was eventually immobilised by the SawTooth or one of the remaining ants. The RedEye spent most of the matches on its side at the edge of the arena, though that did let it survive one round virtually intact. The HailStorm usually got into a firefight with the PressureTank at the start and was the next to be taken out after the erazer.

Overall, hiding seems to be the best way to win, waiting for the enemy to get stuck (or be involved in a friendly fire incident for swarm bots). More usefully, a cannon for long range bombardment is good to have, but close range laser attacks are the most effective.

Regarding HailStorm V ants, accidental death was the main cause of a loss in a complex map as they seemed to have trouble locking onto each other (the ants rely on the target coming to them to some extent). In the open arena, the laser ants were as effective as usual.
(http://i.imgur.com/Vhl877p.jpg)
Title: Re: Autonomous Combat League
Post by: Sleeping_Owl on April 06, 2016, 07:47:37 pm
Lancer (25 parts).
Uses three grenade-lances to charge the opponent.
(http://i.imgur.com/k7KcWeU.png)

https://drive.google.com/open?id=0ByzpYkai1-8lVTZkRFVOdnRMUHc (https://drive.google.com/open?id=0ByzpYkai1-8lVTZkRFVOdnRMUHc)

After a few fights I did it looks quite effective for a lightweight if it manages to get into fight. Some fights ended with it being destroyed from behind (no rear looking camera, no armor on the back) or navigation failure.

Fights:

Title: Re: Autonomous Combat League
Post by: MarvinMan on April 06, 2016, 09:09:02 pm
Cool. That looks like it'll be very effective (even if it is a wonderfully massive amount of overkill against ants). Have you tried it against any of the bigger bots?

The lances look like they make the bot rather long, especially when deployed. It'll be interesting to see how well it handles the tighter or more complex maps. It might be worth adding some elbows to fold the lances and a bit more armour so that it can take on the big bots that the grenades will be most effective against.
Title: Re: Autonomous Combat League
Post by: z26 on April 06, 2016, 09:35:39 pm
unconventional design, i like it.
Title: Re: Autonomous Combat League
Post by: Sleeping_Owl on April 06, 2016, 10:09:07 pm
I tried it on heavier plasma shooters. It is still effective, but you can see it take serious damage. The main problem seems to be lack of rear camera, but it can't have more parts and stay lightweight.
It is quite bad in tight spaces or just fighting near the not smooth wall, but I don't think adding a vulnerable elbow in front of the bot will make its life easier.
If I ever decide to make it middle or heavy I would add rear camera, turret and grenade elevation, and make it less vulnerable to top-down fire (HailStorm can fire over its armor).
Not a conventional turret but something like what was on Lancer's early prototype. Lance arcs would go through armor on bot's body.
(http://i.imgur.com/OGbEb1t.png)



Title: Re: Autonomous Combat League
Post by: MarvinMan on April 06, 2016, 11:37:48 pm
Fair enough regarding not being able to fit any additions into the part count. I had trouble getting a bot into the middleweight class.

As unwieldy as they may be, the tail of grenade lances looks pretty cool. Have you thought about relaxing the distance/proximity requirements for detonating the grenade? It looked like there were a few cases where the grenade passed close enough to do major damage without being detonated. It could be effective to deploy all the lances in a fan pattern when space allows to give it more chances to get one of them into range.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 07, 2016, 04:47:59 am
Not including the Legacy and Unlimited bots, we now have 11 ACL bots! (14 including the Unlimited) Well done guys!
Title: Re: Autonomous Combat League - RockSlide
Post by: PressureLine on April 07, 2016, 12:41:10 pm
New middleweight competitor!

RockSlide

(http://i.imgur.com/AIuZlTZ.jpg)

Weighing in at 35 parts, RockSlide is a middleweight Plasma Shooter capable of tracking and engaging both ground based and aerial opponents.

Get RockSlide here (https://drive.google.com/open?id=0B42Wtrlp8xi9V1FGNk01ZWpUOEE).
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 07, 2016, 02:56:15 pm
The turret on that one seem to work well, and the bot is a lot mare stable than it looks. It is a bit slower than most of the other bots, but that appears to be a good strategy in multi-bot matches as being late to the party means that it only has to deal with mopping up the survivors.

Perhaps now would be a good time to add some new variations on the basic ACL to keep things interesting. Here are a few ideas I've been thinking about:
Sumo/unarmed mode: bots with no conventional (damage causing) weapons try to flip each other or push their opponent out of the arena. Could possibly add some extra environmental hazards.
Air/spacecraft: similar to the current ACL rules, but bots capable of full 3D flight are allowed. Roughly spherical arena defined by distance to a fixed centre point.
Death race: bots try to navigate a track as fast as possible while fending off the competition. No/ limited damaging weapons. Sumo style ramming/flipping attacks encouraged.
Title: Re: Autonomous Combat League
Post by: Sleeping_Owl on April 07, 2016, 07:46:34 pm
Have you thought about relaxing the distance/proximity requirements for detonating the grenade? It looked like there were a few cases where the grenade passed close enough to do major damage without being detonated. It could be effective to deploy all the lances in a fan pattern when space allows to give it more chances to get one of them into range.
I increased detonation distance for fight with bigger bots.
The problem is that Lancer usually damages itself, and with increased distance to do it more often (detonation of almost deployed grenades), so I try to keep distance as low as possible to avoid damage to front wheels (it happened a few times). Also if target part is too close to grenade, Lancer looses target for a short time because there is some space around grenade that is not covered by cameras, these situations can be mistaken for too low detonation distance.

Deploying multiple lances should work great given enough space and some elevation control.

Quote
Perhaps now would be a good time to add some new variations on the basic ACL to keep things interesting.
Sumo - there are infinite forces, infinite masses and partially broken physics. It will need rules to control this.
Air - would be nice. I am not sure if fights in a spherical arena would be easy to watch. My zero gravity big bot construction map (a closed box) is quite unpleasant place, camera is bumping into walls and has weird direction controls.
Death race - it would be nice to get some race challenge running. If I remember correctly, there were only one or two entries in race challenge on old forums. May need separate divisions "flying" and "no jets, no propellers, no mass drives".

ACL as it is now still has a lot of things to try:
Swarms - we still have not seen much teamwork. Ant seems to only have friendly fire protection in targeting system. Same for Nest's missiles.
Unguided grenade weapons - Catapults. Rockets. Spin a grenade, roll it forward, wait, detonate, laser-ram the survivors.
Grabbing devices on laser bots - stick some thin parts into opponents armor. Maybe try to lift his front to attack bottom. Attach a grabber with jet and run jet at full thrust to get opponent out of the map? Maybe not the most practical thing to do but would be fun to look at.
Planes.
Non-plasma hovers - why is plasma the only weapon ever used in the air? Most bots are vulnerable to direct top-down laser attack. Automated bombing is possible in controlled environment, somebody tried it on old forum.

I think the main limit of ACL is that automated combat bots are hard to build and require much time.
Title: Re: Autonomous Combat League
Post by: z26 on April 07, 2016, 08:48:26 pm

Quote
Perhaps now would be a good time to add some new variations on the basic ACL to keep things interesting.
Sumo - there are infinite forces, infinite masses and partially broken physics. It will need rules to control this.
Air - would be nice. I am not sure if fights in a spherical arena would be easy to watch. My zero gravity big bot construction map (a closed box) is quite unpleasant place, camera is bumping into walls and has weird direction controls.

Some rules that could be used
no downforce.
instead of partcount classes weight classes.  If you use mass varying continuums you have to put a cap on their collective value and whatever cap value you chose is whats counted in the mass of your bot.
no hook attaching.

Nonetheless with these rules small hyper dense turtle bots that stalls might reign, so maybe forbidding continuum mass over 1 would be the way to go...  one things for sure unless rules are made to prevent this, all the bots will end up pancake shaped.  Making bumpy terrain so that a compromise between low center of mass and offroad worthiness must be achieved might be the key?

The problem is that the recipie of a good sumo bot is quite simple so this could get dull.  The normal acl reward more diverse tactics.

Also for the camera problems... maybe adding a 0.001 gravity would help?
Title: Re: Autonomous Combat League
Post by: cupid_the_conqueror on April 07, 2016, 09:22:52 pm
I think the main limit of ACL is that automated combat bots are hard to build and require much time.

There ya go! Hit the nail on the head with that one! the very nature of rawbots is slow and tedious. Combat bots take days to make even for  simple ones!

I'm telling ya' pre-programmed behaviors would have saved us TONNES of time (and well, I kinda already do that... just wait for my bot to come out in the next few days! I stole parts from literally every single person on this forum)
Title: Re: Autonomous Combat League
Post by: z26 on April 07, 2016, 09:30:44 pm
In my opinion he reason making bots is so tedious is because of the sensors and programming is too limited, therefore needing to be quite clever to overcome the limitations.  Kinda like how programming done on old hardware use a bunch of tricks to be able to fit in the very limited memory available.

Title: Re: Autonomous Combat League
Post by: MarvinMan on April 07, 2016, 10:06:53 pm
Naturally, sumo bots will need a few rules to prevent unstoppable forces and immovable objects. My main concern that it could easily turn into a less interesting version of the current ACL with the bots just ramming each other head-on until one of them rolls or gets pushed out of the arena.

I had run a very basic test of a spherical combat zone under normal gravity just to see how hard it was to keep a plane in the intended region. Seeing the speeds that the test plane was operating at, it's likely that aerial combat wouldn't work as is would be nearly impossible to hit anything without equipping each bot with a massive amount of fire power.

I dug out the old race track map, and my original bot still works as well as used to. There was one other entry, but I don't know which blueprint it is. This version of ACL should be a lot more feasible now that XFM is standard, as the previous bots were speed limited to prevent the wheels from coming off. The biggest problem I can predict at the moment is bots ignoring the race part and just ramming the opposition into submission at the first opportunity.
Title: Re: Autonomous Combat League
Post by: z26 on April 07, 2016, 10:16:16 pm
the tighter the circle, the slower the bots are forced to move.

battlebots style (non-sumo) real life combat had trouble with too many pushing type bots that were effective but boring to watch.
Title: Re: Autonomous Combat League
Post by: Sleeping_Owl on April 07, 2016, 11:16:33 pm
I have blueprint for my race challenge entry. It was built with XFM installed.
https://drive.google.com/open?id=0ByzpYkai1-8laFFMSGw2UjMxZDQ (https://drive.google.com/open?id=0ByzpYkai1-8laFFMSGw2UjMxZDQ)

A race video I uploaded on old forums is still on youtube.
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 07, 2016, 11:36:46 pm
I don't know whether you'd be able to make the arena small enough to keep speeds down without sacrificing the complex maneuvers that make aerial combat interesting.

Regarding the difficulty of building combat bots, I think most of the people building entries have a couple of basic platforms and turrets that they use as a template to build a particular bots specialised parts around.

I remember that racing bot. It looks like it's probably faster than the later versions of my racing bot that have extra downforce fins.

Re-starting the automated bot racing (Automated Racing League?) seems like the most interesting and viable concept at the moment. Here are the basic rules based on what the last version used:
Aerodynamic downforce only- no jets or propellers
Bots must be limited to a maximum width/length
Mass must be constant
Bots must have the same control continuum/start button setup as for ACL entries
No damage causing weapons
It may be necessary to ban intentional flipping (or ban intentional contact entirely) as it would probably be difficult for flipped bots to know what direction they are meant to be driving around the track.
Title: Re: Autonomous Combat League
Post by: cupid_the_conqueror on April 08, 2016, 12:03:30 am
I have a theory for aerial combat and how it could be done!

Planes would need to have a freeflying mode (where they fly in some pattern or with some form of randomness)

and they would need an attack mode (where they track and attack what they spot)

let them loose on the mid-sized stardust map (the water one? )

see which one finds and destroys the others 1st!  (or just toss 5-8 of Marvins Pidgeons  in the air and see which auto combat flight bot can kill the most)

it would be slow to watch, but sped up / cut it would be interesting! eps if they programmed the flying bots to attack / evade targets on its tail
Title: Re: Autonomous Combat League
Post by: Sleeping_Owl on April 08, 2016, 12:14:29 am
I think most of the people building entries have a couple of basic platforms and turrets that they use as a template to build a particular bots specialised parts around.
It still takes much time. Probably it is worse for me, since I usually try to use new weapon and tactic I never used in ACL.

Here are the basic rules based on what the last version used:
Aerodynamic downforce only- no jets or propellers
Bots must be limited to a maximum width/length
Mass must be constant
Bots must have the same control continuum/start button setup as for ACL entries
No damage causing weapons
Old rules banned jets/propellers completely. Not only for downforce, but also steering, acceleration, whatever else. I think this ban was good and we should keep it.

Why constant mass? To prevent bots from increasing mass before ramming others? Using mass to balance bot on high speed turns didn't really help me, if I remember correctly, side continuums on my race bot were originally added for this.

Maybe we should move this discussion to another thread?
Title: Re: Autonomous Combat League
Post by: cupid_the_conqueror on April 08, 2016, 01:58:14 am
I present to you my 1st AWFUL combat bot! GoblinABRT

if the arena is bumpy, or if the angle of attack is weird, of if its busy trying to turn around. It'll miss no qualms about it! BUT IF IT HITS, GG MF*R

weighs in at 35 parts (I believe, I counted it by hand so I might be off a part or two)
(http://i.imgur.com/xbxRPSH.png?1)

https://dl.dropboxusercontent.com/u/38744980/bp_goblinABRT

for the next one, I think I'll modify the HailStorm chassie as its more stable, and I have more room for hooks to determine when it is and ISN'T okay to fire the rocket (like when climbing a hill!) 
Title: Re: Autonomous Combat League - Blizzard
Post by: PressureLine on April 09, 2016, 01:48:27 am
New middleweight competitor!

Blizzard

(http://i.imgur.com/g07ekIV.jpg)

Weighing in at 49 parts, Blizzard is a middleweight Grenade Shooter capable of tracking and engaging both ground based and aerial opponents.

Get Blizzard here (https://drive.google.com/open?id=0B42Wtrlp8xi9MmwtN2lOWVBIdkE).
Title: Re: Autonomous Combat League - Force Arena
Post by: PressureLine on April 09, 2016, 12:22:49 pm
Just in case anyone missed it, I've completed a new large arena. You can find it here (http://forum.rawbots.club/index.php?topic=61).

(http://i.imgur.com/4oP4ukC.jpg)
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 09, 2016, 08:30:34 pm
The new missile bots seem to be very effective, but appear to have trouble getting the launcher stable and properly aligned. In one case the RedEye got right up next to the Blizzard without being fired at, though it got vaporised as soon as it moved away a bit.

I'll make a new thread for ARL stuff so this one doesn't get too derailed. On a somewhat related note, the navigation system on my automated racing bot works surprisingly well in the bigger ACL maps without any difficult corners.

EDIT: ARL thread is up.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 09, 2016, 10:11:56 pm
Yeah, I made some big changes to the code dealing with launch and detonation timing, which has made it a lot more effective.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 16, 2016, 11:16:54 am
Well, made the version of RockSlide that I had originally intended to make, but it's nothing new really.
Title: Re: Autonomous Combat League - FireBug
Post by: PressureLine on April 21, 2016, 11:32:55 am
New middleweight competitor!

FireBug

(http://i.imgur.com/omYOQnh.jpg)

Weighing in at 50 parts, FireBug is a middleweight Plasma Shooter capable of tracking and engaging both ground based and aerial opponents.

Get FireBug here (https://drive.google.com/open?id=0B42Wtrlp8xi9TzB2eXh5N0FEMTg).
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 21, 2016, 05:53:26 pm
I've had a bit of a play with that and run a couple of 1v1 matches. It works, but it doesn't seem to have enough damage output to match the single turret bots. With higher output turrets or a laser for close range targets it should be very effective.

One thing I noticed was that a lot of the bots have their turrets at the same height, so they are prone to getting disarmed without taking much other damage, especially when the bots get close.
Title: Re: Autonomous Combat League
Post by: PressureLine on April 21, 2016, 10:55:25 pm
Yeah, the single-cannon turrets really don't have enough punch, although in the 1v1's I did it was always around 50/50 on win or lose. I consider FireBug to be more of a test-bed for the targetting and aiming systems anyway, and in that I consider it a success.
Title: Re: Autonomous Combat League - MudSlide
Post by: PressureLine on June 18, 2016, 12:18:06 pm
New middleweight competitor!

MudSlide

(http://i.imgur.com/Y2TCJUS.jpg)

Weighing in at 47 parts, MudSlide is a middleweight Plasma Shooter capable of tracking and engaging both ground based and aerial opponents.

Get MudSlide here (https://drive.google.com/open?id=0B42Wtrlp8xi9N25CQ1NlRzNyNE0).
Title: Re: Autonomous Combat League
Post by: MarvinMan on June 19, 2016, 11:52:32 am
That could be one of the most effective combat bots so far. The armour con tank a lot of damage and the redundancy in the turret helps it stay in the game. The only problem I found with it was that it had a bit of trouble with getting stuck on the small steps where ramp pieces don't fit together properly on corners.
Title: Re: Autonomous Combat League
Post by: PressureLine on June 19, 2016, 12:38:16 pm
Yeah, I'm really happy with it. I really didn't think that the third cannon would make so much difference compared to RockSlide (although the larger chassis helps a bit with stability and such as well) in terms of effectiveness.

And for my next trick:

(http://i.imgur.com/pb1I0UN.jpg)

Building on the relative success of FireBug, my as-yet unnamed triple-turreted beast (no partcount yet either, hopefully it fits within the rules :D [edit: Only just! 75 parts]) is awakening!
Title: Re: Autonomous Combat League
Post by: MarvinMan on June 19, 2016, 12:56:22 pm
What's the ground clearance like on that? It looks like it would have a lot of trouble with ramps.

I wonder if it is more effective to have multiple independent turrets or to aim everything at one target at a time?
Title: Re: Autonomous Combat League
Post by: PressureLine on June 19, 2016, 01:10:24 pm
Ground clearance should be comparable to RockSlide and MudSlide, in the shot posted above the chassis is just sitting on the ground 8)
Title: Re: Autonomous Combat League
Post by: MarvinMan on June 19, 2016, 06:36:17 pm
Cool.

I might have a go at putting a basic collision avoidance system on a quadrotor later, seeing the recent pickup in interest in ACL. Hopefully I can get away with parking directly above the target and raining fire (or grenades) down on it without any real need for aim or a movable turret.
Title: Re: Autonomous Combat League
Post by: PressureLine on June 20, 2016, 08:18:40 am
I've considered doing a rotor-based flyer, but durability is an issue for me.
Title: Re: Autonomous Combat League
Post by: MarvinMan on June 20, 2016, 08:35:25 am
I would have said that a hover bot with rotors can withstand more damage than something similar using jets. I would expect a quadrotor could survive losing an entire rotor assembly, as long as you don't mind it rotating slowly due to the unbalanced torques.

I did build a simple test platform with pitch/roll control to move away from walls, but I think it needs an extra layer of velocity control as it takes quite a large angle to get moving at a reasonable speed.
Title: Re: Autonomous Combat League - Cerberus
Post by: PressureLine on June 25, 2016, 11:51:02 am
New Heavyweight competitor!

Cerberus

(http://i.imgur.com/0v3VW53.jpg)

Weighing in at 75 parts, Cerberus is a heavyweight Plasma Shooter capable of tracking and engaging up to three ground based or aerial opponents simultaneously.

Get Cerberus here (https://drive.google.com/open?id=0B42Wtrlp8xi9bDM2amRTMjFkems).
Title: Re: Autonomous Combat League
Post by: MarvinMan on June 30, 2016, 08:39:15 am
That's an interesting idea, and it seems to have a decent damage output when attacking a target head on with two turrets firing on it. The round armour on this and the mudslide appears to be very effective too.

I think three turrets may have been pushing it to far with the current parts limit as the bot looks a bit wobbly and the turrets are relatively vulnerable. With a bit more armour on the turrets and some bigger wheels, I would expect it could be tough to beat it multi-bot matches.

I'd be interested to see how the cerberus stands up to a 75 part mudslide-XL with 6+ cannons an a single turret.
Title: Re: Autonomous Combat League
Post by: PressureLine on June 30, 2016, 12:17:01 pm
Yeah the wheelbase is probably the bit about Cerberus that I like the least, because of the relative lack of stability. But space did become a bit of an issue as it is pretty cramped inside the armor.
Title: Re: Autonomous Combat League
Post by: z26 on June 30, 2016, 10:57:06 pm
75 parts yet 6 cannons and 3 turrets?
neat.

Edit: Also, i like the color gradient of the armor.
Title: Re: Autonomous Combat League
Post by: PressureLine on July 02, 2016, 10:38:16 am
With a bit more armour on the turrets and some bigger wheels...

I'd be interested to see how the cerberus stands up to a 75 part mudslide-XL with 6+ cannons an a single turret.

I think that I am going to have a go at doing a version of PressureTank with a lower profile chassis and maybe a quad-cannon turret.
Title: Re: Autonomous Combat League - Wheelbot
Post by: PressureLine on July 03, 2016, 10:48:55 am
'New' middleweight competitor!

Wheelbot

(http://i.imgur.com/9eL0QgK.jpg)

Weighing in at 45 parts, Wheelbot is a middleweight Plasma Shooter capable of tracking and engaging both ground based and aerial opponents. This is a 'rules-compliance' edit of my original entry to the 'Automated Bot Arena' challenge on the old rawbots.net forum.

Get Wheelbot here (https://drive.google.com/open?id=0B42Wtrlp8xi9eGQ1WV8wajVhaTA).
Title: Re: Autonomous Combat League - Scylla
Post by: PressureLine on July 10, 2016, 11:51:12 am
New heavyweight competitor!

Scylla

(http://i.imgur.com/57EZXUw.jpg)

Weighing in at 64 parts, Scylla is a heavyweight Plasma Shooter capable of tracking and engaging both ground based and aerial opponents.

Get Scylla here (https://drive.google.com/open?id=0B42Wtrlp8xi9S0YzUkdMaTRwaVk).
Title: Re: Autonomous Combat League
Post by: MarvinMan on July 11, 2016, 11:10:31 pm
Seeing the wheelbot in action again really shows how much the combat bots have improved since the first generation.

The Scylla is an interesting design, and seems to be a good match for the mudslide, although the two bots seem to lose interest in each other after the first encounter. In my small arena, the better aim and higher rate of fire give the mudslide a bit of an advantage.

It'd be fun to add more cannons on each side to take it up to the parts limit.

Have you thought about putting the cannons inside the cylindrical armour section and letting them fire through it? That should keep the cannons fairly central and well protected.
Title: Re: Autonomous Combat League
Post by: PressureLine on July 12, 2016, 08:44:32 am
Currently working on a bot with more of a focus on visual impact. I also have a few other ideas for bots with a less conventional layout, that should keep things interesting.
Title: Re: Autonomous Combat League - Scorpion
Post by: PressureLine on July 17, 2016, 01:34:57 pm
New heavyweight competitor!

Scorpion

(http://i.imgur.com/GRxZaOZ.jpg)

Weighing in at 61 parts, Scorpion is a heavyweight combination Plasma Shooter & Laser Melee bot capable of tracking and engaging both ground based and aerial opponents.

Get Scorpion here (https://drive.google.com/open?id=0B42Wtrlp8xi9R0EteUNkU1MtSFU).
Title: Re: Autonomous Combat League
Post by: MarvinMan on July 17, 2016, 02:23:24 pm
It has it's problems, but that bot is absolutely devastating and looks cool too.

The chassis/navigation system doesn't seem to be performing quite as well as some other bots (in my small map at least), but it was still able to disassemble a mudslide and erazer in a few seconds each. The ramming technique combined with the low body make it particularly effective as it tends to get under other bots and put the lasers through the internal structure.

The elevated turret seems like it should be effective if it got a chance to do an real damage, but it has a tendency to hit and remove the lasers when engaging a target.

How much of the tail structure is necessary? It looks like it might be relatively simple to get the part count down enough to enter the middleweight class.
Title: Re: Autonomous Combat League
Post by: PressureLine on July 17, 2016, 10:13:54 pm
It has it's problems, but that bot is absolutely devastating and looks cool too.

...

How much of the tail structure is necessary? It looks like it might be relatively simple to get the part count down enough to enter the middleweight class.

The main point of Scorpion was mostly 'Rule of Cool (http://tvtropes.org/pmwiki/pmwiki.php/Main/RuleOfCool)' anyway, so the fact that it can easily hold it's own in a heavyweight match is more of a bonus really.

The 4 floaters in the tail section can be removed (was designed that way so that combat damage to the tail wouldn't disable the bot) as can the projector between the cannons. Other than that, without structural changes to the chassis (going single-row rather than double-row) I don't think you could wave a 'middleweight' wand at Scorpion and have it be actually effective. Not that a middleweight Scorpion wouldn't be possible, just that it would need to be built as a middleweight from the start imo.
Title: Re: Autonomous Combat League
Post by: PressureLine on July 25, 2016, 11:48:29 am
(http://i.imgur.com/7gNpgmh.jpg)
(http://i.imgur.com/28zoeBN.jpg)

TIL (https://www.reddit.com/r/todayilearned/): Quadcannon middleweights are pretty tough to beat.
Title: Re: Autonomous Combat League
Post by: MarvinMan on July 25, 2016, 08:51:13 pm
A mutual kill against the scorpion is pretty impressive. I'm guessing the new bot was eventually another victim of the scorpions lasers?

I had been playing with the idea of a a bot armed with a flipper to disable or disorientate its opponents. The basic concept seems to work, but packaging all the autonomous control systems was more of a problem than normal.
Title: Re: Autonomous Combat League - Babel
Post by: PressureLine on July 27, 2016, 01:14:51 pm
New middleweight competitor!

Babel

(http://i.imgur.com/lQTGn9C.jpg)

Weighing in at 50 parts, Babel is a middleweight Plasma Shooter capable of tracking and engaging both ground based and aerial opponents.

Get Babel here (https://drive.google.com/open?id=0B42Wtrlp8xi9dmk3Q1JNbjE1ak0).
Title: Re: Autonomous Combat League
Post by: MarvinMan on April 05, 2017, 09:23:11 pm
I had a go at reworking the broadsides reversible drive & navigation system. The result is fast and doesn't get stuck, but unfortunately it's not really well suited to having a turret and armour fitted.

It is however fast enough that it nearly doesn't need armour, so it should still be useful as a target bot to test turrets against.

On an unrelated note, I'd forgotten how much code was in the turret of my original combat bot. It's probably not going to be fun trying to untangle and modify it to use the XFM inverse trig functions.

(http://i.imgur.com/kXatqxz.png) (http://imgur.com/kXatqxz)

Blueprint (https://www.dropbox.com/s/oy7x6aolp9u2ud7/bp_broadside2?dl=0)