Six Lessons for Unmanned Systems at Wildfires

Posted by Robin Murphy on Aug 24, 2018 at 11:00 am America/Chicago

1. If you have a small UAS, don’t fly it unless you are really, truly authorized to fly. Large wildland fires almost always have a TFR (temporary flight restriction) on all manned and unmanned aircraft- the equivalent of  police yellow Do Not Enter  tape. The TFR makes it safer for the manned aviation which is flying at low altitudes to evacuate people, drop fire suppressants, and monitor the fire. No small unmanned aerial system can fly, even under hobbyists rules, without a) explicit permission from the agency managing the TFR and  b) approval by the FAA. While a small UAS may not physically destroy a helicopter, a collision or a sudden swerve can lead to an accident, so flying an unauthorized small UAS during an event is a lot like dropping small rocks from an overpass onto the interstate— you probably won’t damage the car engine but it still might lead to a wreck.
2. Flying in a TFR zone is feasible. As our experiences at Hurricane Harvey show, it is actually easier to coordinate flying in a hurricane with a TFR. As long as you have an Air Boss, that is the director of Air Operations- it is an official role in the incident command hierarchy,  who is knowledgable about both manned and unmanned aircraft, then it is fairly simple to partition by mission, altitude, area, and contingencies. A TFR helps thin out the non-essential drones, though a few people will still continue to fly theirs illegally. A TRF can allow sUAS to fly above the 400 ceiling.
3.  Night flying is a big win for authorized small UAS. Recent events as well as numerous tabletop exercises sponsored by AUVSI and CRASAR with CAL FIRE, Department of the Interior, and other agencies including those that supply smokejumpers, continue to highlight that drones are particularly useful at night when the fire crews are down and manned aviation does not fly. As seen in our deployment to Hawaii, getting permission to fly at 1,000 ft AGL gives much more coverage than 400 ft and there is less wireless dropout.
4.  Mapping is useful but the key isn’t flying, it is determining what kind of mapping is needed and how the data will be processed (and by whom)? This merits talking about in depth.
   • What type of mapping is needed? There is quite a bit of difference in flying, processing, and comprehension.
   • Do many decision makers need to easily comprehend the situation? Then perhaps Hangar 360 which gives a 360 panorama that is easy to understand (versus a birds eye nadir view) and move around over a website and only requires sending URL is the way to go. Do the GIS people need updated maps for analytics? Then photogrammetric processing of orthomosaics or digital elevation maps may be in order .
   • What is the area to be mapped- 10 acres? 100? or 10,000? Large areas mean higher altitude, longer range assets. An Institu ScanEagle or a manned aircraft can cover more and more effectively that trying to piece together a quilt of areas flown by a Mavic- and that assumes that the Mavic team could reach the locations needed to map those quilt pieces. Small UAS flying at low altitudes are great for mapping small areas of interest or hot spots but not good for larger areas.
     • Don’t just fly missions, think through the mapping process- don’t just fly and then drop off an unsorted SD card (yes that happened). Put each mission on a separate folder or better yet separate SD cards! Don’t mix mapping and recon— devote a flight to recon with imagery and video, then fly a separate flight for mapping What post-processing will be used? Datamapper? Pix4D? Agisoft? Does the agency have that license or a way to pay for it?
5.  SUAS may provide better image quality for local areas or help with small fires, but remember that large fires are supported by manned aircraft, higher altitude UAS, and satellite imagery.  sUAS may be able to fly under the smoke plumes that interfere with imagery from high altitude assets. Infrared for large wildfires is already provide by manned aircraft, higher altitude UAS, and satellites. For example, agencies have been using thermal systems that orbit an area and determine the boundary of the fire for several years. And again, sUAS may be valuable for hot spots and small fires.
6. Don’t forget about ground vehicles and the types of imagery they can produce. An autonomous ground vehicle with a ground-based system such as GroundVu that captures 3D imagery can provide rapid damage assessment (think of it as disaster StreetView). Justin Adams, our president, and GroundVu CEO Myles Sutherland just returned from assisting with the Shasta and Redding fires. They collected over 100 miles of 360 degree imagery that embeds in ESRI products in just a few days. The citizens of Fort Bend County loved GroundVu- they could click on the path down their street and see the state of their house.