Resolve: lowering voltage limits to 9V will significantly reduce robot speed. Show me the data!

I’m pulling this discussion from the 2022 draft rules thread to initiate a debate-like discussion.

The concern was raised by @bukajlag that “By reducing voltage to only 9V there would be no reduction of speed because you can buy motors for 9V with the same parameters as for 12V but with a higher current…”

Power is indeed a product of voltage and current (P = IV) and we do not want to impose a rule that has little/no effect. However, there are questions as to if these equivalent high current motors exist/are practical, if a typical battery setup be able to discharge enough current to drive these motors, etc.

This is not a discussion on whether or not the rule is warranted, if there is a better way to reduce/maximize speed, etc. This is just a matter of resolving the motion: lowering voltage limits to 9V will significantly reduce robot speed. At the very least, it may be a worthwhile rule change if it effectively caps the speed at what the fastest robots are now capable of.

Please back your agreement or disagreement with the motion with examples and/or data.

Thank you for those that bought on the discussion and in advance for those that will add to it here.

As I tried earlier in a paper to describe the rulemaking practices in RoboCup Junior., I must reiterate that for any provision of any rule, we must first examine our objectives of that rule.
For example, in this discussion, we must first answer this question:
Why do we want the robots to walk more slowly?
We probably want the technical and learning aspects of students to take precedence over the financial and facilities. In other words, we want the team that has made the most “effort” to “learn” to be superior to the team that has the most “facilities”. If this is true, we are moving exactly against the goal. For example, in this case, the 9-volt limit only allows teams that are more financially capable to buy better, higher-quality motors. While 12 and 15-volt motors have more common and are cheaper and more accessible. (Higher current motors need better drivers.) This means that the voltage-limiting action will result in teams that do not have access to expensive 9-volt motors(and high current drivers) being virtually eliminated or severely weakened by full-featured teams. In such a way that they have nothing to say in the competition.
I do not know how understandable this is for you but If you feel this does not make sense to you, it’s because this is my experience in a country like Iran, which faces severe restrictions on parts supplies.

@javadrah please do not misdirect the point of this thread; I’m hoping we’re all after the same goals. To clarify:

We find the need to consider an effective and practical means of limiting the speed of the robots; hence these discussions. One such consideration is reducing the voltage (as automated refs and fuses are deemed far less practical even though they may be more effective). This thread is not about whether or not we should be looking into limiting the speed but merely about if we were to consider a rule change to limit speed for whatever reason, would a voltage limit be a good solution?

A question came up as to whether a voltage drop from 12V to 9V would even been an effective way to limit speed (even if limiting it to the current top speed we’ve seen). If someone can demonstrate that a voltage drop of 3V would not be effective, then we can take the voltage drop proposal out of the equation. There are so many factors to consider in untangling what ‘effective’ is - higher current motors options may have less torque, be more expensive, put too much drain on the batteries, be less available, etc. Then, of course, there is the ability to over-volt a motor so there is only so far motor spec sheets will take you.

If the proposal to drop the voltage to 5V was being made, then the effect of this drop on the maximum robot speed may not be a discussion at all!

Higher quality components/materials/tools/etc will always exists and some teams will always have easier access to them than others - that is not a discussion for this thread although is a very big concern; a discussion of whether or not voltage drops may increase the “money game” effect could take place. Right now, we can avoid having to go into that with what is (hopefully) a much more pragmatic question as to the effectiveness of the proposed voltage drop in the first place. Even if it served to give a speed limitation of what the robots currently are performing at, then it may be a good rule to consider.

@javadrah I’d like to take a look at your paper - could you please message it to me?

I’ll respond to @stiebel s arguments from Draft rules for RoboCupJunior Soccer 2022 released - #43 by stiebel on a technical level here:

The intention is to make perception and planning more important and better motors less important in gameplay.

9V is meant to not outlaw Lego bots from new teams with their AAs - we know we’d effectively limit most teams to 2S.

The thinking behind 2S is the following: For motors to have the same torque they need the same current, to get the same current at lower voltages they need gearboxes (weight penalty) or lower inductance motors (a.k.a. fewer windings / lower kv in DC/BLDC terms). Motors with lower inductance generate more back-emf as rpm increase. So teams would have to make a trade-off between stall torque, max (usable) rpm and weight: to keep the exact same power (i.e. significantly higher currents) heavier motors would become necessary. All this should - and we have as far as I can recall only heard assertions that teams would buy different motors but have not seen data to contradict this reasoning - where I looked at motors (which I admit was just HobbyKing in this instance) the tradeoffs I described broadly hold true - also losses from ohmic resistance increase, heat dissipation in the motor controller increases and so on - lots of things that should add up to a reduction in speeds (we think).

But if you can demonstrate how voltage/current don’t matter here’s another though: Would going the other way with voltage at least help level the playing field if slowing down robots isn’t going to happen (i.e. allow cheap motors to go close to the same speed as expensive ones)? Like if something up to 48V (or 36V or 24V) nominal battery voltage were to be allowed allowed (48V nominal is the maximum before high voltage safety rules apply)?

Cheers David

Hello forum,
I will present some of our soccer-bot details to explain why reducing the voltage is a bad idea in my opinion:
We started driving our robots with the polo 12V MP Motor:

this is a 12 V DC Motor with a stall current of almost 2A, so we are talking about max 24Watt per motor.
With a 4mot-bot we could reach 100Watt for movement.
This was an easy setup and we had very high speeds, but we needed to give 25-30% power to get initial torque and so there was no chance to have precision with low speed.

Then we decided to change our strategy from brutal force to precision and we switched to a brushless motor with hall sensors

our setup is the 12V Version with a 16:1 planetary drive.

This motor has a load current of less then 400mA and a rated power of 3.3Watt and because we saw that a high current when stalling brings no advantage we programmed our motor drivers to have a 1A max current.
So even if we would use the highest possible power that would be 16V times 1A times 4 Motors and that would be 64 Watt.
But we are fine with just 20W alltogether as our robot runs 2hours on a 4S-3000mA Li-Io setting including running all the computing stuff as well.

But this setup is only possible with industry style drivers and these start working at 10V.
As mentioned above by javadrah 12V Motors are commonly used and our setup is 30$ per gearmotor and 35$ per driver.
We did not find anything similar for 7.2V.

But what we find is this:

the same sized motor as we started, but with a 6W - 6A rating.
combined with this:
Pololu G2 High-Power Motor Driver 24v21 which provides 21A continues current on 2S Lipo without even getting close to be hot we can easily have 7V*7A * 4 Motors → 200 Watt for motion without having heavier equipment than the stuff we started.
But then again we will face the same “torgue at low speed” - problem.

So in my opinion reducing the voltage will NOT decrease the speed of the robots but will decrease the possibility to go with high precision.
And that means we will push the whole competition into the wrong direction.

So lets face it (and Mike made that clear!): we want to reduce the speed of the robots to gain precision.
And I really appreciate this, as we decided to go this way three years ago already.
But reducing the voltage will not lead to this goal.

If a power restriction is the way to go, then we have to restrict the POWER and not the voltage.
I must admit that it is way harder to control this than just checking the nominal battery voltage.

Measuring the speed during the game is difficult but possible. (using a mobile phone with slow motion cam could be used to stop the time over a given distance) It could be used if in doubt and not during the full game. And we are not talking about exceeding the limit with 10% but massively.

But then - why would we want to decrease the speed?
We want to see higher precision!

I have an easy solution:

We reduced the size of the robot from 22cm to 18cm.
Why not reducing the goal width from 60cm to 50cm.
At the moment the strategy “fast to the ball and kick as fast as possible” will goal with a chance of 45% (goal to field ratio is 60/132)
If we reduce the goal width to the same robot/goal ratio we had before (22/60 → 18/50) that would decrease the chance to 37%.
And this would be near to zero effort for rule change.
But this will force teams to turn from pure force to precision.
And then again: reducing the robot size further down to 17cm or even 16cm with reducing the goal to 40cm will definitely produce smaller and slower robots that will have to improve precision.

I think 18cm and a golf ball as introduced for 2022 will show this effect already.
Going further into this direction would work far better than any technical restrictions that are hard to control.

Sorry for this long post, but I think we all have to think about future robots and games in a holistic view and not in a discussion about the voltage of the batteries.

Best regards
Roland

1 Like

Hi Roland,

If I have your selections correct:
Original: 12 V, 1.8 A stall, 800 rpm, 1.7 kg.cm torque, 82 g mass
9V solution: 6 V, 6.0 A stall, 1000 rpm, 2.3 kg.cm torque, 82 g mass

That indeed looks like the only tradeoff seems to be current and you’ll get more RPM at higher torque. That’s assuming, however, that we’re following the recommended voltage. There is nothing saying that you can’t run the 6V motor at 15V as long as you’re willing to replace them sooner rather than later! That’s the issue I ran into when just comparing motor specs.

As discussed, we’ve considered speed tests and cameras (I’ve ran some simple experiments as you suggested with calculating speed from video) but introducing official recordings to the matches is not practical at the moment. Remember we’re talking about what regulations need to be followed internationally, on strict time schedules, and with limited volunteers. Anyone who’s ran a competition before will attest to the headache that may be induced if we require that every match be recorded and have to be analyzed the day of competition!

Your argument for precision and decreasing the goal width is very interesting! What about making it too easy for a goalie to defend the smaller goal though?

We hadn’t considered field modifications before and have added that to the list of considerations. Thank you for that! At this point, I believe you’re correct that we have some time to study current matches to determine how big the problem really is and what the potential solutions are; we may be at the speed limits of the bots anyway before they’re just tearing up the fields.

Also thank you for pointing out that brute force may not be better than precision on the field - good for teams to hear that.

Mike

Like others said, we would just drive with, for example 5Amps instead of 2.
So we would have no speed reductions, and it would just more work to find Lipos with 9V.

Regards
Robotronic

For completeness:

All references to the 9V limit have been removed from the 2023 rules draft (see this thread and the GitHub generally for current drafts).

Regards
David

1 Like

Final rules are out now and the 9V limit isn’t in there and there are no plans to put it in in the future.