After 3-4 anxious days of living on the edge of four bushfire zones (Update a new fire has flared overnight 12km from my home) and seeing helicopters and water bombers flying over to Richmond air base, it seems to me that our fire fighters could use some help from the local drone community.
The NSW RFS has done an awesome job in what was arguably the most difficult week we have faced in some years, but seeing these planes fly past at 10-15 minute intervals I got to thinking life would be a lot easier if the Rural Fire Service had access to a fleet of drones.
Turns out Drones have been used very successfully in California to monitor wildfires earlier this year. Using MQ-1 Predator UAVs normally used for military and security purposes the drones were piloted from hundreds of Km away from the site. With infrared and high definition video and the ability to stream this video and data back to base the drones are very effective at surveillance and spotting new outbreaks, they are also reasonably immune to human error that comes with flying in thick smoke and having to use instruments (something which not every pilot is equipped to handle in a bush fire situation)
Think about this, a long-range drone such as the Predator can stay in the air for nearly a whole day. Most humans in a single-seater aircraft would be lucky to do 4-8 hours (I think this is probably a stretch). Assuming you wanted 24 * & coverage of NSW you could arrange 4-5 of these Drones to fly patterns every few hours like a big loop over the affected areas.
Humans in Aircraft are Expensive
The thing that makes aircraft expensive to build and run and limits their range is humans.
When you put a human up into the air, you need a stack of systems and extra hardware to carry the human and to keep them alive which increases weight and reduces the range.
If you build a Drone there is no cockpit, seats, instrument panels, air conditioning, canopy, control sticks and other equipment to operate the aircraft, a very large chunk of the equipment needed on a manned aircraft is eliminated when it is unmanned.
Every time you remove weight you increase range for a given power and lift.
And frankly humans can’t perform effectively for 22 hours and in the dark they are unlikely to see much and on surveillance duties they are unlikely to add much value.
Have a look at this footage captured by a group called Cividrones.com (the name makes me think they are ex defence types, no info can be found online about them except their twitter account).
This amazing footage of a quadcopter drone fitted with a Gopro flying through burnt out buildings shows how versatile drones can be in these situations.
Drones are much cheaper to run and build, you can put them together in weeks rather than years it takes to get aircraft built, they can be set to do certain tasks autonomously such as fly search patterns to observe for fires or to keep station on a particular point.
I know of teams of University students in University of Sydney, UNSW and societies of hackers that could assemble a world-beating set of drones in a very short period of time.
Potential Methods of Operation
If we had a spotter drones with a hyperspectral or infrared cameras, they could conduct surveillance on a very wide range and keep flying 24*7 at a fraction of the cost of manned aircraft.
Some of these UAV using various types of high-tech cameras are able to ascertain different types of plants, one University researcher I know can tell from Satellite photos if Power Stations are running at full steam or if a paddock is fully irrigated or if you have weeds.
I they can do this, then spotting fires from 10,000ft should be relatively easy. Also if the military drone builders can work out how to drop a bomb on a single building then waterbombing a fire front should also be achievable.
When the spotter drone finds fire outbreaks, they could alert the operation centre and provide live video, in this instance a drone QuadCopter could be launched from the roof of a truck and go look over the ridge or into the valley to see where the outbreak is.
A water bomber drone could be launched and deployed to the same co-ordinates within 30 minutes of the outbreak and not wait until it has taken hold.
The aircraft aspect is relatively easy, most of it is 20-year-old technology, the harder part is the software and systems to manage this, the semi or fully autonomous drone is essentially a flying software play and we have the guys sitting in Sydney that can do it.
With autopilot and mission planning software from a company like 3D Robotics a UAV company founded by Chris Anderson the former editor of Wired.com and an aircraft designed for maximum time aloft such as the Zephyr which managed to smash the record for unmanned flight by recording a time of 14 days without landing due to its solar power and lightweigh construction. This is probably not quite robust enough (as the winds hit 50kmh out of my window) but a high aspect ratio solar powered drone could easily staff aloft for days at a time with operators sitting on the ground in safety.
I think the biggest challenges are both human and operational. The hardware and technology is available to do this. Both RFS and CASA (Civil Aviation Safety Authority) have come out warning drones owners to stay away from fires or risk a $30,000 penalty.
I can understand their issue, its hard enough trying to manage air traffic control in a fire zone without a bunch of drones doing sightseeing trips however I think the RFS with CASA’s blessing needs to engage with the local drone community and see if they can contribute their skills to help.
There is no reason that they couldn’t dovetail their capabilities into a normal air traffic control framework, in fact the a surveillance drone could become the virtual Control Tower for a whole operation.
It’s obviously not the time now, but in a few weeks when the fires have settled, I challenge the drone community of Sydney to come together and build a fleet of piloted and autonomous drones that can help fire fighters with surveillance, close in support, water bombing and mission control and I challenge NSW Rural Fire Service and CASA to come together and work out how to facilitate the introduction of drones into active service.
Here is my dream team of guys to take up the challenge.
- The University of Sydney Field Center for Robotics http://www.acfr.usyd.edu.au/ These guys are amazing, they build UAV and all manner of robots, they have run projects using UAV to find different types of weeds.
- University of NSW Robotics http://www.robotics.unsw.edu.au/
- Professor Luiping Wang from RMIT http://scholar.google.com.au/citations?user=xoGSfpEAAAAJ&hl=en Expert in control systems and autonoumous flight
- Mechatronics Engineer Fernando Vega – Robotics expert
- Cameron Cooke – Geo Image Analysis and big data expert
- UNSW Create – Big team of mad keen electronics and design students – Rapid prototyping and construction http://createunsw.com.au/index.php
Its my contention these guys could build a working surveillance drone capable of 12 hours of continuous flight within 6 months that gives command and control capability to NSW RFS and then after that start working on other drone roles such as water-bombing and locally deployable quadcopter for tactical intelligence for local brigade commanders could be developed to work with the surveillance drones.
I think it’s about time we created our own technology for firefighting drones and not wait for the US to let us have it, also much of the technology which drones use very successfully in the US Military for surveillance and targeting is is most likely classified information so if we want the best methods we should be creating it ourselves.
If you think someone should be on this list or you think that I missed something, please let me know.