Questions and remarks from Ascend NTNU

Questions about the competition rules

Moderator: IARCadmin

Post Reply
brage
Posts: 1
Joined: Tue Jan 31, 2017 2:41 pm

Questions and remarks from Ascend NTNU

Post by brage » Wed Feb 01, 2017 7:51 am

As there is no general section in this forum, this post will be posted in Rules Interpretation although some aspects of this post will go beyond questions about the rules. This post serves as a small review with follow up questions concerning the next event. The questions and remarks will be enumerated to easily talk about them.

  1. Rules Interpretation and judgement
    1. A note at the end of the rules specifies practice times to be the day before the competition day. This year however there was no practice times. As we need to deconstruct and reconstruct our UAV due to travel, a window of time to test that all systems are functional would be welcome. If there is no opportunity for any testflights, it would be nice if the note is removed to remove any uncertainty, though we believe giving all teams a possibility to test their systems before the competitions would raise the overall level of performance in the competition.
    2. As the pattern of the floor is now fixed it would be nice if the pattern was easily available on the website, and the specific pattern mentioned in the rules. If some teams does not know they can get the pattern (this affect new teams especially) the teams with the pattern have an unfair advantage. And yes, we do have the pattern.
    3. Under the “Assembling the Hardware for an Obstacle Robot” section of “Information About the Ground Robots”, the PVC pipes are said to be between 1 and 2m. The rules however only specifies <2m, and one of the obstacles in the competition was indeed shorter than 1m. It would be nice if the instructions for assembling the obstacle robots were updated not to imply that all obstacles are >1m or changes the rules to fit the assembling manual.
    4. The top plates of the obstacle robots were flapping back and forth due to the propellers. Will the top plates be fastened from now on? A small tape or equivalent seems to do the trick.
      The judgement of Elegance of design and craftsmanship and Safety of design to bystanders (part E and G respectively) ought to be judged not only by the symposium presentation but also by a physical inspection of the UAVs. By requiring everyone to bring their UAV to the symposium, one could easily check all the hardware feats of the drone, and as a bonus, everyone can see each other’s UAV and discuss the different solutions and so on. One possibility would be to hold a small event right after the symposium where all the teams can display their drone for inspection by the judges and answer questions. This could also be a opportunity for the teams to get to know each other before the competition and inspect each other's drones.
    5. During the competition it was hard to figure out what each team were trying to achieve during their attempts. An announcement before each attempt as to what the purpose of the attempt is, and some way to signal autonomous/piloted mode would greatly increase the audience’s ability to understand what's going on, and therefore increase the enjoyment one get from watching.
    6. Though we are grateful for receiving the award Best Performance American Venue, it would be nice to know why we got the award. We got the 3rd highest static score, so knowing what bumped us up to first would help us do even better next time.
  2. Safety at the arena
    1. Looking at the videos from the Asia/Pacific venue there are constraints placed around the arena to keep everyone at a safe distance. We would very much like to see such constraints in the American venue as well.
    2. Safety Testing the drones. In addition to the safety switch, a test of airworthiness of the UAV would be nice. Something simple like a piloted flight across the arena and back again would suffice.
    3. Accident may happen regardless of safety measure, and as such first aid kits and trained personnel should be present at the competition.
  3. Next years competition
    1. Dates for the competition. Travel expenses is by far the greatest single expense in our budget. The earlier we get the competition dates the better. This also goes for summer internships and work in general. When can we expect to know the dates of the 2017 competition?
    2. We liked the idea of having one hotel for the teams very much. What hotel is not to important right now, but knowing whether it will happen or not is, as if there is not a shared hotel we would need to find accommodations elsewhere.
All in all we at Ascend NTNU liked our experiences at IARC and look forward to once again meet everyone and compete this summer. We do not expect all the questions to be answered right away, but from experience we know that these things take time, so better to be too early than too late. If other teams agrees or disagrees with these points, we would love to hear from you, in addition to the officials at IARC.

User avatar
IARC Organizer
Posts: 17
Joined: Mon Jul 27, 2009 3:26 pm
Location: Jasper, GA, USA
Contact:

Re: Questions and remarks from Ascend NTNU

Post by IARC Organizer » Wed Feb 01, 2017 2:12 pm

1a. It is our intent to offer a self-moderated time for teams to assemble and test their vehicles/systems in the actual arena prior to the event. Sometimes our access to the Georgia Tech facilities have not been as promised when we contracted for the space. We are at the mercy of those who open up the buildings (in the past this has been the GT Athletic Department). We are in the process of negotiating a move to a larger facility on the Georgia Tech campus for 2017, and the behind-the-scenes logistics will probably be different (better). “Self-moderated” means that the teams coordinate among themselves for flight times and durations. The intent is that teams use any time provided prior to the event to assemble and align their systems. All too often teams come unprepared to the IARC and do last minute “development” which can consume arena time doing things that they should have done at home. Teams should monitor the situation and stick to an arena-use schedule that is mutually agreeable. The Organizers will not mediate these testing periods other than to negotiate with the university for the practice time slot.

1b. The Official Rules state that “The floor space inside the arena will be optically textured with an unknown random or regular pattern. “ (see page 11). The Organizers have worked to assure that the floor patterns used at the American and Asia/Pacific Venues will be the same to eliminate any unfair advantage, but everyone has the same opportunity to photograph the floor pattern on their first visit to the IARC (as many have done). In some cases, those requesting a digital photograph of the floor pattern have been supplied with such in advance. Therefore all teams are on an equal footing in terms of accessing knowledge about the floor pattern as well as any other physical feature of the arena. No promises are made in the Official Rules to provide flooring specifics, so teams not requesting photographs of the flooring pattern from either the Asia/Pacific or American Venues, from another team, or from the Organizers, by their inaction, put themselves at a disadvantage. Please note that the the Official Rules do not guarantee a particular floor pattern, only that “The floor space inside the arena will be optically textured with an unknown random or regular pattern”, so it is entirely possible that the Organizers could change the floor pattern from year to year. Gearing a system to key-in on a particular floor pattern could be putting a team at risk as we would remain entirely compliant so long as the Asia/Pacific or American Venues maintain an optically textured random or regular pattern.

1c. Go with the pipe length being 1<length<2. We will check to see if one of our pipes was cut short and replace it.

1d. The top plates are being redesigned and the modification will be posted shortly. Problems with the top plates, as they are, fall into three categories. First, the suggestion that a “floor tile” material be used (which is what we used in 2016) proved to be problematic because the floor tile squares turned out to NOT be dimensionally stable. They sagged with time so that they were not totally flat. They will all be replaced with thin aluminum (approx 1/16 inch thick). The second problem as observed, is that the downwash caused the plates to flip up. A fix was to put a piece of ribbon or string on the plate to prevent it from flipping higher than the length of the string. We are looking into a different method of restraining the plate. Also, there is the potential that an incredibly strong downwash could push the plate so hard that it compresses the spring and actuates the microswitch prematurely or accidentally. For this reason, we are going to increase the stiffness of the recommended spring so that any aerial robot making physical contact will activate the switch, but the likelihood of a downwash activation would be less likely. The third issue is with the starting of the iRobot Creates. The on/off switch is occluded beneath the plate. When the plate is constrained in its flapping angle, a special hooked tool had to be used to activate the on/off switch. We will be putting a small hole in the plate to allow this switch to be accessed with a straight tool from the top.

The Judges base their “Elegance of Design” determination on the actual machines as well as the Symposium description. That determination is not made solely at the Symposium. The “Mixer” held prior to the competition day is intended as the time for teams to get to know each other. Bringing their aerial robots to that event is entirely acceptable. Also this interaction can occur during any practice time. One problem is the scheduling of facilities availability. The IARC occurs during the time of year when new students are on campus for orientation. Therefore we are vying for the same spaces that are used for orientation, so scheduling large areas for team displays is another issue.

1e. The Judges agree that we need an announcer. The Asia/Pacific Venue has a full-time announcer and the American Venue needs to do the same. Everyone should know which team is up (although we had that on a screen in 2016), what they are going to attempt, when they “go autonomous”, and when the run ends.

1f. The rationale for the granting of awards is determined by the Judges. They are available after the Team Awards Banquet to answer questions pertaining to performance improvements and what strengths and weaknesses they see in the teams.

2a. One reason to move to a larger venue on the Georgia Tech campus is to eliminate crowding around the arena. The Asia/Pacific Venue also uses line monitors with aluminum poles to knock down air vehicles that go too far out of bounds. While there are kill switches, this is an added layer of safety. During the 2016 Asia/Pacific Venue, no less than two aerial robots had poles thrust through their propellers to knock them out of the after going out of bounds.

2b. So noted.

2c. Two large first aid kits have already been purchased for the American Venue in 2017. We will no longer rely solely on the first aid provided by the Georgia Tech facility in which we are operating. These kits will be at the side of the arena.

3a. Two thing drive the competition dates. Availability of the facilities and coordination between the American and Asia/Pacific Venues. The American Venue is currently being negotiated for 2017, but the university does not schedule events farther out than several months. The IARC would like to schedule the space a year in advance, but the university will not do that. We hope to announce the American Venue dates some time in March when the calendar for the desired space opens for July/August. Then we must also coordinate with the Asia/Pacific Venue in an attempt to get the venues in sync and nearly simultaneous.

3b. In the past we have attempted to get “room blocks” at a specific hotel to get better rates for the teams and so the teams could stay close together and close to the venue. Unfortunately, when we negotiated the room block, teams often rejected the room block to stay at another location, stay with local friends or family, or cancelled. This causes trouble with the hotel, and we must “guarantee” the room block, paying for unused rooms. This is too much risk for our budget, so we leave the choice of hotel up to the individual teams. We have also tried to get dormitory rooms for the teams on campus, but because the IARC occurs during the university’s orientation time, all of the dormitory rooms are allocated to newly enrolled students who are coming for their orientation week.

amiller27
Posts: 1
Joined: Mon Feb 06, 2017 9:31 am

Re: Questions and remarks from Ascend NTNU

Post by amiller27 » Mon Feb 06, 2017 8:34 pm

I have a couple questions I'd like clarification on; one of them is based on Ascend's earlier question, which is why I'm posting this as a reply here.

Regarding their question 1b, a previous answer on this forum stated that "the floor textures at both the American and Asia/Pacific Venues are the same and will continue to be used throughout Mission 7." However, the above post states that "it is entirely possible that the Organizers could change the floor pattern from year to year." Does this mean that we are to ignore the previous answer? If so, would it be possible to have a picture of next year's floor pattern?

And a separate question: in the official rules, page 21, the section "Target Ground Robots" is marked with a note that says "These specifications are “typical”, and are provided so that teams can test realistically at home. During the IARC runs, initial conditions may vary to induce more random behavior in the ground robots." Are there any restrictions on how the conditions are allowed to "vary?" For example, are all of the numbers in the "Target Ground Robots" section allowed to vary? If so, could we have bounds on their allowed values? In addition, is it possible that the source code for the iRobot Creates (other than those numbers) will be different on the competition robots from what is posted on the website?

User avatar
IARC Organizer
Posts: 17
Joined: Mon Jul 27, 2009 3:26 pm
Location: Jasper, GA, USA
Contact:

Re: Questions and remarks from Ascend NTNU

Post by IARC Organizer » Wed Feb 08, 2017 2:04 am

Question 1b stated that, "teams with the pattern have an unfair advantage". Our response points out that the Official Rules do not guarantee any pattern ever. We stand by that, but practically speaking, we have coordinated the pattern to be the same between both venues and we plan to reuse the same arena flooring for several years for economic reasons. So the flooring is not likely to change, BUT were both venues to decide to change the flooring pattern for some reason, the Official Rules allow that to occur without notice.

Regarding the Target Ground Robots, the code that is posted is the code that we use. The conditions that may vary to induce more random behavior has to do with the Creates themselves and physical conditions. Specifically, we have found that when Creates are programmed identically (for example the circle radius of the obstacle robots), variations in the robots results in slightly different behavior (in the case of the obstacle robots, they all navigate circles of slightly different radii even though they are programmed identically. This is probably due to wheel friction as some robots carry heavier pipes (longer pipes), or wheels become coated with dust from people walking in the arena, or there are variations in the robots' mechanics due to factory tolerances). We really don't know all of the sources of these variations, but since we are seeking random behaviors, these variations help us to make the arena nondeterministic. The Aerial Robots must accommodate these variances, and that is a behavior that we are trying to foster.

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest

cron