Announcement

Collapse
No announcement yet.

Task Ideas Released (And open for YOUR review!)

Collapse
This topic is closed.
X
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Task Ideas Released (And open for YOUR review!)

    In case you haven't heard, it's time to mark your calendars for the 2022 Maritime RobotX Challenge on November 11-17, 2022! Although 2022 may seem like a long way off, there's never enough time to prepare. So, gather your teams (virtually, of course) and review the Task Ideas document! Make your feedback count and let us know what you think here on this thread.

    Access the Task Ideas here: https://robotx.org/2022/#resources

  • #2
    The UAV safety inspection requirements seem a little light, and will likely need some additional requirements such as prescribing and testing geofencing and geofence breach behaviour, prescribing and testing loss of datalink behaviour, and other safety related constraints like maximum size, weight, speed etc. There probably needs to be some discussion regarding positioning systems (ie gnss vs visual vs beacons vs fused etc) and how design choices in that domain impact safety cases.

    Comment


    • #3
      Thank you RobotX organizers for this early release of draft tasks. We are excited to compete in Sydney in 2022. Our initial comments on the tasks are below. Please note, we are not against any of them, but on first reading some initial comments/concerns popped up in relation to the UAV-based tasks. Also we are wondering why there are no sonar/underwater based tasks (e.g. detect an underwater object or follow a pipeline etc). Apologies everyone for the size of this post.

      Whilst there is an obvious “coolness” in coordinating ASVs and UAVs, apart from Task 6 we could argue that a UAV is not required at all to complete the tasks. Additionally, we have some concerns about the risk/benefit of integration without more details that may make it unattractive (at least for us) to attempt particularly Task 3 with a hyperspectral camera.

      Some initial points/ideas we think all teams would need more detail on are (in no specific order) are:
      • Hyperspectral cameras are quite expensive in relation to student competition budgets. Additionally, the UAVs used to carry these may start to exceed 2kg. I know that the insurance cover for our hyperspectral camera would not cover installation on a smaller drone over water. Can the wildlife simulants be made such that they look authentic so we could use standard RGB cameras and cheaper micro-UAVs?
      • We know the UAV Outback Challenge successfully uses drones with international teams, but will RobotX organizers be working with CASA to get necessary exemptions or permits for all local and international teams to operate non-recreationally in the competition area?
      • When we read this, we are assuming we can’t use the “sport and recreation use” clause for drone use. Therefore, will teams be required to have RPA operator accreditation and an RePL?
      • Will the teams have to get extra insurances for the use of drones in this competition?
      • It appears that the top of the courses are only about 100m from restricted airspace. Will our drones be required to be tethered? What happens if a UAV crashes into the water? Will they be required to be waterproof and/or have floatation on them for recovery?
      TASK 2
      ---------
      • Love the idea of a navigation task, but can this be done without a UAV if a team chooses ? That is, can we just use the ASV’s on-board sensors to complete this task? These onboard sensors could also be used to provide a map of the channel and obstacles.
      TASK 3
      ---------
      • We like the idea of the task but think the use of a hyperspectral camera may be restrictive to some teams. Will the RobotX organizers be providing teams with a low-cost hyperspectral camera or have a supplier provide them at an attractive rate? If so, will you also be providing teams with copies of the wildlife simulants or drawings with instructions where to get the paints/coatings used? If not, can the organizers make the simulants realistic such that RGB cameras can be used and provide copies or instructions for teams to make their own for testing?
      • As per Task 2, given the size of this course, it could also likely be completed using sensors onboard the WAM-V. Could teams be given the option to do this without the need for a drone?
      • If a hyperspectral camera and UAV is required, could the organizers possibly provide an image from their drone for the teams to use if they wish (possibly with a points penalty) for teams that do not have the budgets or capability required (or crash their drone and camera during the competition)?
      TASK 4 – A classic, no comments.
      TASK 5 – A classic, no comments.
      TASK 6
      ---------
      Given the very little details, we can only speculate some comments.
      • Will the colored disks be magnetic?
      • Will the colored disks be light weight so micro-UAVs could carry them?
      • Can the UAV remain/land at the helipad or must it drop/release it? Where must the UAV land?
      • We foresee that there may be segmentation issues of putting the colored disks on top of the detect and deliver element as the UAV could also see the sides which will have colored disks and other shapes on them. Also if a large lid is placed on the top it may then cast shadows over the walls making it hard for some teams. Would it be better to have a small floating platform in the coarse that has the disks on them? This could also be safer for the organizers rather than them climbing on the element to place them on top.
      • The UAV light beacon can have colors that may clash with the safety, remote control and autonomous mode light systems required for the WAM-Vs. We assume that the beacon will not use the standard red, blue, yellow and green colors used by teams on their systems particularly when all courses are being used?

      Comment


      • #4
        Thanks to all the RobotX organizers for planning all these details thus far. We (Team Bumblebee) hope to compete in RobotX 2022. Here are some comments that we have for the proposed task ideas:
        1. Is there any specific reason the AUV component of the competition was removed completely? We were hoping to use RobotX as a platform to develop solutions for some of these technologically difficult problems involving ASVs and AUVs.
        2. We would like to request for the AUV component of the RobotX tasks to be added to the task list for RobotX2022.
        3. This might sound like something obvious, but we would like to clarify that entire control of the drone has to be autonomous, just like the Surface Vessel with no external human communication.
        4. Are drones absolutely necessary if the tasks can be done by the Autonomous Surface Vessel system alone (with the exception of the drone manipulation task which requires a drone) ?
        5. Is the UAV retrieving the ball that had been fired by the Surface Vessel? Or will there always be a ball at the task that the UAV can pickup? And are we allowed to modify the ball's properties / structure to make it easier to be retrieved by the UAV?
        6. We understand that the dates for the competition have been finalized. However, would it be possible to shift the date of the competition (after consultation with all teams) to December (perhaps 9 Dec - 15 Dec) instead, as Nov is typically the exam period of our University?

        We look forward to your replies. Thanks!
        Last edited by hashir.zahir; 02-13-2021, 11:58 PM.

        Comment


        • #5
          This is great feedback so far, everyone! Our RobotX staff will get back with some answers/questions shortly.


          Quick answer to to your #6 question, hashir.zahir:


          We understand that the dates for the competition have been finalized. However, would it be possible to shift the date of the competition (after consultation with all teams) to December (perhaps 9 Dec - 15 Dec) instead, as Nov is typically the exam period of our University?
          The dates that were decided were based off the availability of our venue. We wish there was a way to find a set of dates that didn't conflict with each team's exam schedule. Do the current dates (11 Nov - 17 Nov) cause a huge conflict for your team? In 2016 and 2018 we have provided an exam room on-site that students were able to complete their exam remotely. Would this accommodation help some of your exam conflicts?


          In the meantime, keep the comments coming! What questions do you have about the proposed tasks? Any other suggestions or requests?

          Comment


          • #6
            Task 3 is as follows:
            The Wildlife Encounter and Avoid task requires the UAV to identify objects of interest and inform the WAM-V to circumnavigate the object. These objects of interest will represent Australian Marine Life such as platypi, turtles and crocodiles and will each be painted in distinct coatings with unique spectral signatures to enable identification and classification with a hyperspectral camera.

            Issue: The use of hyperspectral camera may cause extra spending and extra effort. Is it possible to amend it to the use of normal color camera?

            Comment


            • #7
              Hi Julianna, has there been any update on the queries? Looking forward to the reply!

              Comment


              • #8
                hashir.zahir So sorry for the delayed response! But, I come with some answers (or soon to be answers). Keep an eye out this week for more information on RobotX 2022!

                Comment

                Working...
                X