I have a question regarding tactical lack of progresses. Depending on where the checkpoints are placed, it might be possible for teams to
get to a victim on a floating wall if the checkpoint aligns to a floating wall
get the exit bonus
even though the robot is actually not capable of getting this scoring (that means if the checkpoint were somewhere else, they would only be able to get the score randomly, by luck).
Of course, one could argue that the first point can be resolved by proper arena design by simply not aligning checkpoints to floating walls. Maybe it can be considered to mention this in the rules. But how about the second point? Let’s say the starting tile has three walls and one opening in one direction and a team has a simple mapping algorithm, but during the run the map is resetted (after a LOP) and the robot loses the starting position. Now the robot reaches the starting tile at the end of the run. Normally, the robot would just continue with the maze, but if the team does a LOP at this point, the robot will be resetted to the starting tile and think it already has visited the only neighboring tile it can visit and will therefore stop at the checkpoint, scoring the exit bonus.
I am wondering if something like this is considered giving the robot information about the maze? The rules don’t say anything regarding tactical LOP and the team does not give this kind of information explicitly to the robot (just implicitly by doing one/multiple LOP at the right time). Furthermore this problem can’t be resolved with a proper design of the arena because if the starting tile has only one wall, the team could just do three lack of progresses (visit a different tile after every LOP) and after the third LOP, after all the neighboring tiles were visited, stop again at the checkpoint and get the exit bonus.
Of course one could argue that the team knows their robot very well, but it’s not the point of the rules to get the scoring for cases like that.
Also basically all LOPs are tactical, but the difference is that the team kind of gives information to the maze with these two mentioned kind of LOPs.
So the question is how to handle the first and especially the second case? I’d say avoiding 1. is task of the designer of the arena, but 2. can only be avoided partly (having less than three walls next to the starting tile), so teams should not get the scoring if they obviously do multiple LOPs to get the exit bonus as this should be considered giving information of the maze to the robot. Of course this might be difficult to recognize for inexperienced referees again.
What is the opinion of teams and the TC on this?
Best,
Jan
EDIT: I made a drawing to visualize what I mean. I assume that the robot the team uses is capable of creating a map and always visits the neighboring unvisited tiles. If all neighboring unvisited times are visited, it will stop because it thinks it is done and because no route planning was implemented. This way it will get the exit bonus without having anything illegal in the software, just by doing two LOPs at the right time. I think this should not be allowed.
Hi Jan,
In the score chapter, we write about the exit bonus as follows.
3.5 Scoring
3.5.11 Successful Exit Bonus. A successful exit bonus is awarded when a robot successfully finishes a round on the start tile. It needs to stay there at least 10 seconds (this is to simulate the retrieval of a robot from the disaster zone). Each victim successfully identified will be awarded 10 points (see 3.5.1)
In the Summary, it is written as follows, and in End of play it is defined as follows.
Summary
If the robot can find its way back to the beginning of the maze after exploring the whole maze it will receive an exit bonus. The robot will also earn a reliability bonus if the robot can exit the maze with a minimum number of restarts. If the robot can find its way back to the beginning after exploring the maze it will receive 10 bonus points per identified victim as an exit bonus.
3.7 End of Play
3.7.2 The round ends when: i) the time expires. j) the team captain calls end of round. k) the robot returns to the start tile and gets the exit bonus.
Therefore, the robot must return from the other tile to the Starting tile to get the exit bonus.
However, as you say, it is better to write that it will clearly return from other tiles in the Scoring part.
However, since 2018 rule has already been submitted to GC, I would like to include it in 2019 rule.
the problem is something else: If a team does a few LOPs and because of the LOPs the robot stops at the starting tile, does this count as giving information of the maze to the robot? So if the team would not have made the LOPs at this point, the robot would not have stopped.
2.1.1 Robots must be controlled autonomously. The use of a remote control, manual control, or passing information (by sensors, cables, wirelessly, etc.) to the robot is not allowed.
I think that doing that is a violation of this rule.
Hi Jan,
In situations in your figure the team declares LoP twice. At the second restart, did the team behave exactly the same as the first one?
If same, can you prove that the team gave information to the robot?
Did you check the team’s program ?
It is very difficult and sensitive.
It is easy to doubt, but we can not just judge the information given to us.
the team definitely has no modifications in the program. The team actually asked me if this is okay Now I am asking this here to clarify. The problem is that the algorithm the team uses just works that way. It is not preprogrammed, like “if three LOP then stop”, but the algorithm will think that it finished the maze and therefore stop. Is it okay to do multiple LOPs because the team knows that the robot will in that corner case stop at the starting tile? I don’t think it is okay, but I am not sure.
Basically the team knows that the algorithm will stop the robot in this particular situation. So the team exploits this knowledge by doing three LOPs, because the team knows that the algorithm will think that it finished the maze.
The team asked the question to you.
I’m misunderstanding, you want to judge the team. Sorry.
Okay, I will try to clear about “exit” end “return to the Starting tile”.
What is “exit”.
The robot must return to the Starting tile and stay for at least 10 seconds.
What is “return to the Starting tile”.
The robot will come back to the Starting tile from other tiles by itself.
It is not a “return” that the team puts the robot at the he Starting tile.
For example, in the following cases, the robot has not succeeded “exit”.
The robot failed during scoring run. The team declared LoP. The last checkpoint tile is the Starting tie. The team placed a robot on the Starting tie. The team restarted the robot but the robot does not move. And 10 seconds passed.
The team started scoring run. However, the robot continues to rotate on the starting tile. The team declared LoP. The team restarted with the robot, but it will not move. 10 seconds have elapsed.
＊There may be other examples, The important is that the robot has to “return to the Starting tile”.
It will clearly state in the relevant item number (3.5.11) with the rule of 2019.
Similar to what Jan is asking, our robot is programmed that if it lack of progresses 3 times at the same place, it will return home. If we don’t force the robot to return home by manually triggering LOPs, is this allowed?
Now I am asking this here to clarify. The problem is that the algorithm the team uses just works that way. It is not preprogrammed, like “if three LOP then stop”, but the algorithm will think that it finished the maze and therefore stop. Is it okay to do multiple LOPs because the team knows that the robot will in that corner case stop at the starting tile? I don’t think it is okay, but I am not sure.