Earlier this week (posted June 12, 2009), Ron Zeeman had a very nice discussion at a training meeting for Utah County Search and Rescue. He identified several coordination and logistical constraints that constrain how we approach near-term work to maximize our chance of having a real impact. Here are the constraints

  • Any personnel participating in a search must be trained and sign the liability waiver from the Sheriff's office.
  • The command post (CP) s often too crowded for current searches, making it seem unlikely that we'd be able to have a large team and trailer at the CP.
    • Would a command post have a generator that we could tap into for UAV controls? –James
  • CPs may not be easy to access, meaning that we need to minimize how much infrastructure we require.
  • Trained search personnel are rare, so we need to minimize how many are required to participate in a UAV-enabled search.

Proposed Plan

  • Because of liability and training issues, the quickest route to deployment is to train Ron Zeeman and another searcher on how to prepare, deploy the UAV, start a search plan, and use the video analysis software to do search.
  • Because of space issues, we need to create a minimal configuration that can fit into the back of a pick-up or into a couple of backpacks.
  • Create other configurations that scale to more equipment and personnel resources.

Scalable Configurations

  • Minimal Configuration: Ideally, this would require only two people and one laptop. We could probably get by with two laptops. The laptop(s) would run the software that controls the UAV and the software that supports the video analyst. A single operator should be able to do both tasks. A timeline would look something like the following:

(1) Prepare equipment, power, and antennae for launch.

(2) Launch the UAV: fully automated launch is ideal, but having one person on RC while the other starts the operator interface is fine. **You may want to consider a bungee launch. I haven't yet seen one, but from what I've discovered about them, bungees are much safer and more consistent. –James

(3) Input a search plan, and begin execution.

(4) Switch to the analyst's interface. Include mosaicking, scrubbing, zooming, brightness control, etc. Include a monitoring window in the analyst's interface that actively reports information on the UAV's flight – time to completion, battery life, height above ground, etc.

(5) Tie the operator and analyst's interface together so that, for example, when the analyst role wants to scrub, the UAV enters a loiter pattern.

(6) Land the UAV, plug in new batteries, and redeploy.

(7) Evaluate video offline using more sophisticated tools like seeability and automated target detection.

Equipment required for the minimal configuration includes:

* Cables
* Antennae
* Spare batteries
* Laptop
* Power (batteries for equipment of small generator)
* RC equipment
* Wonderbox?

  • Optimal Configuration: This is the configuration that we are trying to reach right now. It is optimal with respect to minimizing the expected time to locate the missing person. It allows a full-scale deployment of MAGICC Lab's trailer and a full set of analysts, IC, operator, and ground searchers. The idea is that this would be useful for deploying at the CP.
  • Constrained Configuration: This should be somewhere between the minimal and optimal configuration. The idea is that the trailer or truck could be parked close to the place where the UAV would launch, and then equipment and personnel deployed to the desired launch location, perhaps as much as a mile away. The launch point could be different from the location of the place where video is analyzed and the UAV is controlled. For example, the MAGICC Lab's trailer could be parked at a location, the UAV launched a mile away, and then the search is controlled remotely.
    • Is there any reason that the UAV control and the video analysis must be at the same place? Would it be of any use to separate the two? –James

Other Constraints

  • We need to solve the problem of how to satisfy FAA requirements.
  • We might need to consider one of us being trained as a searcher.
    • This is something I'm interested in anyway; if it needs to be done I am happy to do it. –Spengy 13:00, 12 June 2009 (EDT)
wisar/a-plan-for-moving-forward.txt · Last modified: 2014/08/11 19:43 by tmburdge
Back to top
CC Attribution-Share Alike 4.0 International = chi`s home Valid CSS Driven by DokuWiki do yourself a favour and use a real browser - get firefox!! Recent changes RSS feed Valid XHTML 1.0