Poll

What do you like doing most in RA2/RA3/robot combat games?

Building robots
Local single player battles with manually-controlled robots
Local multiplayer battles  with manually-controlled robots (PvP with controllers or a shared keyboard)
AI-only tournament battles
Other?  Please comment below.

Author Topic: Robot Rumble 2.0 - Robot Combat Simulator - Under Development  (Read 187743 times)

Offline kix

  • RR2 dev
  • *
  • Posts: 3452
  • Rep: -3
  • H
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1600 on: January 28, 2020, 05:35:08 PM »
Idk if the fix will fix this, but small parts tend to clip into opponent. Lets say a small spinner tooth hits an opponents, the tooth will most likely get stuck and it will be that way until it breaks off

Offline cjbruce

  • Super Heavyweight
  • Posts: 963
  • Rep: 11
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1601 on: January 28, 2020, 07:10:31 PM »
Idk if the fix will fix this, but small parts tend to clip into opponent. Lets say a small spinner tooth hits an opponents, the tooth will most likely get stuck and it will be that way until it breaks off

Nope.  That is something that happens with compound colliders and discrete physics like we are using.  I tried switching to continuous physics a few times, but it made wheels bounce all over the place so I switched back.

The blur cylinder completely eliminates the problem if we set it to turn on at a low enough speed.  Maybe I can play with this a bit.  It will only work for spinner weapons though.

Offline kix

  • RR2 dev
  • *
  • Posts: 3452
  • Rep: -3
  • H
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1602 on: January 29, 2020, 03:52:33 AM »
Idk if the fix will fix this, but small parts tend to clip into opponent. Lets say a small spinner tooth hits an opponents, the tooth will most likely get stuck and it will be that way until it breaks off

Nope.  That is something that happens with compound colliders and discrete physics like we are using.  I tried switching to continuous physics a few times, but it made wheels bounce all over the place so I switched back.

The blur cylinder completely eliminates the problem if we set it to turn on at a low enough speed.  Maybe I can play with this a bit.  It will only work for spinner weapons though.
Well spinners pretty much cause the issues

Now there is an issue where if a part reaches 0hp and breaks off it will get stuck in the bot and cause the bot to gyro and not drive in general. How about when the part reaches 0hp and breaks off, it disables collision, launces upwards and then regains collision while in air.

Offline cjbruce

  • Super Heavyweight
  • Posts: 963
  • Rep: 11
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1603 on: January 29, 2020, 12:23:05 PM »
The 29January2020 Alpha build is up!

https://robot-rumble.itch.io/builds

This one fixes the crashing bug introduced in the Bleeding Edge Build.

It also adds a 12V AmpFlow A28-150 motor.   This is a less-powerful lower voltage motor we use for our lightweight class robots.

The COMTrackers are no longer visible in this build.

Offline CyarSkirata

  • Antweight
  • Posts: 114
  • Rep: 0
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1604 on: January 29, 2020, 01:20:54 PM »
I haven't said anything about it until now since it slipped my mind, but I've been having an interesting situation with Panic Attack these last few builds.

The slow-weapon version no longer gets phantom gyro until the main body of the srimech comes off the gearbox, but every version still has that issue from jan 17 of having too much traction to steer properly.

Also: Is it possible to turn off the pink COM spears of doom on detached parts?
Stuff I Done Competed In:

Offline cjbruce

  • Super Heavyweight
  • Posts: 963
  • Rep: 11
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1605 on: January 29, 2020, 01:38:45 PM »
I haven't said anything about it until now since it slipped my mind, but I've been having an interesting situation with Panic Attack these last few builds.

...every version still has that issue from jan 17 of having too much traction to steer properly.

This effect happens in real life too.  4-wheel tank steering can be problematic.  With 2 wheels, the wheels only have to roll forward or backward when turning.  Whenever you have four wheels with some wheels in front of others the wheels have to skid sideways.  The longer the wheelbase compared to the width, the more pronounced this affect.  The better the traction, the more pronounced the effect.

I believe that the solution for Panic Attack is to shorten and widen the wheelbase as much as possible, though I haven't tried.

I'm planning to set the Wheel Mass Ratio at 10-15% for future builds.  Panic Attack drives best when I set it at 1%, but this makes other robots completely undriveable as the wheels skitter all over the place and aren't capable of pushing the chassis.

Also: Is it possible to turn off the pink COM spears of doom on detached parts?

Done!  Check out the Alpha build I just posted.

Offline kix

  • RR2 dev
  • *
  • Posts: 3452
  • Rep: -3
  • H
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1606 on: January 29, 2020, 04:39:51 PM »
It also adds a 12V AmpFlow A28-150 motor.   This is a less-powerful lower voltage motor we use for our lightweight class robots.
Ok but when are we getting brushless motors?

Offline cjbruce

  • Super Heavyweight
  • Posts: 963
  • Rep: 11
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1607 on: January 29, 2020, 04:43:42 PM »
We’ve got one in the art pipeline.  I haven’t decided how to handle the performance characteristics.  Maybe it is only usable with a belt drive?  Maybe it has very little low speed torque?  Has anyone built a real robot with brushless motors?  A heavyweight?

Offline kix

  • RR2 dev
  • *
  • Posts: 3452
  • Rep: -3
  • H
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1608 on: January 29, 2020, 04:58:30 PM »
We’ve got one in the art pipeline.  I haven’t decided how to handle the performance characteristics.  Maybe it is only usable with a belt drive?  Maybe it has very little low speed torque?  Has anyone built a real robot with brushless motors?  A heavyweight?
There are many HW bots with Brushless motors, Pulsar/Magnetar is one (iirc it used brushless motors for drive). Atm the issue is that many compact bots have to rely on weaker motors for their spinners, where bigger bots just can use the montenegry(etek) motors.
For the compensation for the motor itself to not be overused. Well IRL as you possibly know, Brushless motors are more fragile and are tend to overheat easily. Because the game atm has no overheating. Maybe compensate with the motor being heavy (unlike irl where its lighter than brushed), till we get overheating stuff.

Offline Team Lightning

  • Addicted to combat robotics, good at annoying people, and wanting to get to -10000 rep XD.
  • Beetleweight
  • Posts: 168
  • Rep: -4
  • Education is important, robots are importanter.
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1609 on: January 29, 2020, 07:04:08 PM »
We’ve got one in the art pipeline.  I haven’t decided how to handle the performance characteristics.  Maybe it is only usable with a belt drive?  Maybe it has very little low speed torque?  Has anyone built a real robot with brushless motors?  A heavyweight?
There are many HW bots with Brushless motors, Pulsar/Magnetar is one (iirc it used brushless motors for drive). Atm the issue is that many compact bots have to rely on weaker motors for their spinners, where bigger bots just can use the montenegry(etek) motors.
For the compensation for the motor itself to not be overused. Well IRL as you possibly know, Brushless motors are more fragile and are tend to overheat easily. Because the game atm has no overheating. Maybe compensate with the motor being heavy (unlike irl where its lighter than brushed), till we get overheating stuff.
For overheating, couldn't you add a variable that is incremented whenever the motor is on but not spinning?


Offline BATTLEMASTER

  • Antweight
  • Posts: 24
  • Rep: 0
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1610 on: January 29, 2020, 07:52:07 PM »
My Tombstonish horizontal spinner is much easier to control with the COG correction :claping  Actually, I find all my bots much easier to control!

Totally looking forward to the new damage system so I don't have to go crazy building ultra-destructable bots :dance:

 
Bot destroyed.jpg

Offline CyarSkirata

  • Antweight
  • Posts: 114
  • Rep: 0
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1611 on: January 30, 2020, 01:34:53 AM »
I believe that the solution for Panic Attack is to shorten and widen the wheelbase as much as possible, though I haven't tried.

The issue I have with that is twofold. Partly that there isn't really room to do that without clipping until multiple wheels can be belted to one motor, partly that the real thing - albeit with six wheels - had a wheelbase just as long as this one and handled fine.
Stuff I Done Competed In:

Offline WhamettNuht

  • *
  • Posts: 1302
  • Rep: 12
  • Robot Building Drag Queen
    • View Profile
    • Awards
  • Discord: WhamettNuht #1457
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1612 on: January 30, 2020, 03:06:45 AM »
The issue I have with that is twofold. Partly that there isn't really room to do that without clipping until multiple wheels can be belted to one motor, partly that the real thing - albeit with six wheels - had a wheelbase just as long as this one and handled fine.

Have you tried building the robot with 6 wheels? Having two wheels in the middle of the robot will most likely give Panic Attack more control when turning as it has more of a 'pivot point'.

If you're concerned about weight, you could always have the front pair of wheels motor'ed and the back ones mounted to a free axle?
Damn I should probably put something fancy in this bit huh?

Offline Wambo

  • Beetleweight
  • Posts: 126
  • Rep: 0
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1613 on: January 30, 2020, 04:01:46 AM »
I think, I found an issue with the new build. Some asymmetrical spinners, that were balanced before are now unbalanced

Offline cjbruce

  • Super Heavyweight
  • Posts: 963
  • Rep: 11
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1614 on: January 30, 2020, 06:05:24 AM »
I think, I found an issue with the new build. Some asymmetrical spinners, that were balanced before are now unbalanced

The COM calculation has changed several times over the past few months.

Method #1 - The first iteration weighted ALL colliders based on their volume.  Even ones that shouldn't have been included (i.e. invisible colliders that were used as attachment points for the Botlab).

Method #2 - The next few iterations removed the invisible colliders, but assumed all colliders had the same material density.  This means that selecting "steel" or "aluminum" for a collider farther out didn't shift the COM.

Method #3 - One of the most recent iterations attempted to determine the COM based on the location and mass of each component (i.e. the entire shape) rather than using the individual colliders.  This is the desired method, but due to a bug in the calculation it did not correctly account for non-uniform scaling of the component.  In the case of Panic Attack, Method #3 placed the srimech's COM about 1.5 meters out in front of the robot.  This caused the srimech to respond too slowly.  Also, if broken off, the srimech would hang vertically from a point in the air.

For the current 29January Alpha Build, I have reverted to Method #2 above.  It isn't what I want, but at least it doesn't sometimes place COMs 1 meter outside of the robot.

I would really like to get Method #3 working correctly.  In the meanwhile, just be mentally prepared to tweak the balance of asymmetrical spinners over the next few builds.

Offline cjbruce

  • Super Heavyweight
  • Posts: 963
  • Rep: 11
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1615 on: January 30, 2020, 06:06:20 AM »
I believe that the solution for Panic Attack is to shorten and widen the wheelbase as much as possible, though I haven't tried.

The issue I have with that is twofold. Partly that there isn't really room to do that without clipping until multiple wheels can be belted to one motor, partly that the real thing - albeit with six wheels - had a wheelbase just as long as this one and handled fine.

Maybe we can play around with friction coefficient to see if we can better match reality?

I sense another slider in our future! :)


Offline kix

  • RR2 dev
  • *
  • Posts: 3452
  • Rep: -3
  • H
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1616 on: January 30, 2020, 07:28:17 AM »
Oh yeah i forgot to ask this back yesterday, but when are we getting elastics? Something to retract arms of flippers

Offline cjbruce

  • Super Heavyweight
  • Posts: 963
  • Rep: 11
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1617 on: January 30, 2020, 08:10:45 AM »
I believe that the solution for Panic Attack is to shorten and widen the wheelbase as much as possible, though I haven't tried.

The issue I have with that is twofold. Partly that there isn't really room to do that without clipping until multiple wheels can be belted to one motor, partly that the real thing - albeit with six wheels - had a wheelbase just as long as this one and handled fine.

Maybe we can play around with friction coefficient to see if we can better match reality?

I sense another slider in our future! :)

Another thing to consider is that the real life Panic Attack wasn't driven by someone using a keyboard.  It was driven with a very expensive controller with high precision analog sticks (i.e. not the relatively inexpensive ones that come in an Xbox or PS4 controller).  This enabled the driver to carefully control the signal to the wheels.

Panic Attack turns fairly quickly with the keyboard if you don't try to drive forward or backward at the same time as you turn.  I am hoping to get a Spektrum WS2000 USB receiver to pair with our Spektrum DX6 radio transmitters.  This will allow me to test the RR2 driving dynamics with the actual controllers we use in real life.  I am curious to put Panic Attack through a series of tests to see how the simulated robot performs compares to the real robot.

Does anyone know if Panic Attack's creator is on GTM? :)

Offline cjbruce

  • Super Heavyweight
  • Posts: 963
  • Rep: 11
    • View Profile
    • Awards
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1618 on: January 30, 2020, 08:12:30 AM »
Likewise, does anyone have any real-life experience with 4-wheeled robots?  It would be nice to get some turning and driving feel feedback.

All of our robots use 2 wheel drive.

Offline WhamettNuht

  • *
  • Posts: 1302
  • Rep: 12
  • Robot Building Drag Queen
    • View Profile
    • Awards
  • Discord: WhamettNuht #1457
Re: Robot Rumble 2.0 - Robot Combat Simulator - Under Development
« Reply #1619 on: January 30, 2020, 09:57:45 AM »
Likewise, does anyone have any real-life experience with 4-wheeled robots?  It would be nice to get some turning and driving feel feedback.

All of our robots use 2 wheel drive.

Our former featherweight was four wheeled skid steering.

Feel free to email me about this :)
Damn I should probably put something fancy in this bit huh?