Hi forum,
we just had the the old discussion again:
Teams are using IR distance sensors.
The open and lightweight rules are absolutely clear : “the robot must not emit infrared light”.
But then there comes the next sentence telling that IR distance sensors are allowed as long as they do not distract the other robot.
I do not like this. It means arguments on the field whether the distraction is high or low and it often
leads to “why do you not use IR-Filter on your camera”
So here is my suggestion:
in lightweight we eliminate the permission to use IR-Sensors at all.
Teams can use ultrasonic sensors.
It is much better to know this from the very beginning than being told during the match that the sensors are not allowed.
And in open we cancel the whole chapter.
In open we can expect the teams to use IR filters on their cameras.
There we have to be aware of colors and IR light should not be banned at all.
If you all agree, we could change this right now at least for open, as this rule change does not restrict any robot that is already built, but prevents unnecessary discussions. So this could be a rule change even for 2020.
We also recently had the problem that you mention.
I would say that it might be better to just ban the use of 870nm - 1000nm band (the ir ball uses 920nm - 960nm). This way the teams can still use many ir distance sensors and lidars and there cannot be any interference.
This is in my mind a better option since it does not limit students in their choice as much. Lidars for example have smaller beam width than ultrasonic sensors which can come in handy for localization, which seems emphasized in the new rules.
Hi NesTo,
you are absolutely right, lidars have much smaller beams than ultrasonic sensors and can be a much better choice for positioning matters. However, banning a specific bandwidth will even increase the opportunities for arguments at the playing field. What if for example only 10% of the intensity is at 980nm and the rest above 1000nm. Is this allowed or not? How will you check this on the table?
This might be a very German characteristic of mine, but I rather have strict and clear rules than discussions at the game.
But again, I would suggest opening IR totally for 2vs2open. (The bandwidth-sensor is the human eye and everyone can see whether it is visible light or not) and ban IR completely for lightweight.
This would be easy to control and would even sharpens the difference between the two leagues.
Greetings, Roland
Hi stibel!
The absence of unnecessary bans, this has always attracted me in RoboCup. It is not necessary to prohibit all IR sensors for LW, so that the ban does not apply to passive IR sensors or to modified IR sensors with IR LEDs turned off. In my opinion, only active IR radiation should be prohibited for LW. During field disputes, IR radiation is easily checked through the smartphone’s camera.
Greetings, Sergei Kosachenko
On the other hand, I witnessed when the lighting in a sports building was an IR interference to a team with passive IR sensors without a frequency filter, and this team tried to convince the organizers to turn off all the lights in the building during the games, referring to the rules. Therefore, if the rules mention the ban on IR radiation, then you need to mention the need to use frequency filters on IR sensors. This is a dilemma)
Hi Sergei,
I totally agree with you that passive sensors should stay allowed. And the presence of IR Light from the building should be considered as “background radiation”.
But if you have a look at this and this, you see what I mean by banning IR-sensors from lightweight. These sensors produce a light that is much brighter than the IR ball and cause problems again and again.
As long as we have the “however, optical sensors(e.g.infrared-distance sensors) may be used as long as they do not affect other robots.” in rule 3.2.2 the discussions will go on. It is not necessary to use these sensors and I rather ban them in the first place than telling a team they can not use them, when they spent lots of time working with them before and depend on these sensors.
This year marks the 10th anniversary of the introduction of the pulsed IR ball. Back in 2010, we witnessed a stunning display by German team, T’n’T, as they dismantled teams with grace that few teams have ever demonstrated, and few have matched since. This standard was adopted as it was “more robust to interfering lights” (according to 2020 rules) and “enables the sensors to easily distinguish the infrared light emitted by the ball from infrared emissions from other sources” (Elekit). The performance of great teams who understand this move have confirmed that this is indeed the case. This allows them to utilise cutting edge sensors, such as LiDARs (including those operating on different frequencies) to allow the competition to progress. Let’s not undo the progress that has been made a decade ago and embrace LiDARs. After all, it is clear that interference is only experienced by teams which are unable to accurately demodulate the signal. To remind ourselves about what Robocup is about, I would like to suggest Prof Manuela’s amazing speech at the symposium in Sydney (https://youtu.be/rSO0nYAsO2c). Looking forward to more innovation, and for some teams, rediscovering innovations from a decade ago
For me it seems that there is almost no difference in difficulty between ir ball detection and colored ball detection. Image sensors such as Pixy camera and OpenMV reduce the difficulty substantially, they were not around when the league started, but they are now. The time needed to have something working is almost the same, but the difference is that in ir when you get that working you are basically done, there are many more ideas which you can explore with vision and there is much more knowledge to be gained. Ir balls are also expensive, need batteries and produce problems with interference, you can just color part of a sensor that is orange but you cannot prevent it from using ir light to function.
