Drone trial


Frank Pope - Chief Operations Officer, STE

Date Published

It was 7pm on the 28th May, and around two dozen wardens of differing seniority had assembled at Oliphants Reserve in the ANPR west of Kruger National Park to witness the trial of conservation’s most-hyped new technology: drones. The moon was not yet up, and when a young bull elephant wandered past (as if to inspect his purported saviour) he was only spotted by the lights of one of the ten or so 4WD vehicles that were lined up on the grass.

Chris Meiser, an ex-USAF officer once in charge of weaponising small UAVs (drones) for the Air Force and who now runs a business building and selling “Falcon “UAVs, pulls out a black bag the size of a quiver of golf clubs. He unzips it and starts pulling out sections of white wing and throwing them to the ground (with slightly exaggerated vigour as if to demonstrate their rugged construction).

Less than five minutes later he has plugged the three wing sections together and clicked them into place on top of a shoebox-sized pod that carries the batteries, the camera payload and a parachute for landing. He slides in a tube that has a motor the size of a salt-shaker on the end with a folding propeller in the front, and another that has a V-tail rudder-elevator combination in the back, and proclaims the vehicle ready.

A few pre-flight checks – all twitching servos and a brief snarl from the motor – and he’s walking backwards with the drone stretching a long line of elastic that is attached to the bull-bar of a warden’s Land Rover. He lets it go. With a woosh the thing is up and climbing to its initial target altitude of 150 metres, on full autopilot.

The waypoints and altitudes are all programmed by the operator prior to the flight, using a laptop that connects to the drone via an antenna which (in this case) is placed on top of the support vehicle. The drone’s flashing light (a booklight attached to one of the wings, for our viewing convenience) quickly disappears and we’re left with the view from the on-board infrared camera.

There are two cameras on this ‘mission’. One is a daylight camera, the other infrared. The daylight one shows the lights from the vehicles below as star-like pinpricks. Switch to the infrared camera and the whole drone-view screen (a separate display) glows in eerie, Iraq/Afghanistan shades of grey. The grass around our vehicles glows with reflected heat, and the people are visible as spots of white.

As the drone flies way silently and invisibly to its first waypoint at the far end of the airstrip we’re all crowded around the car to see what it’s seeing. Chris is letting the Falcon fly itself but is using the PlayStation-style controller to pan and tilt the camera to look around. A big white shape appears to the left. A rock still pulsing with the sun’s heat? Then it moves. An elephant. The camera tilts. Three more appear in front.

The Falcon costs $20,000, with another $3,000 or so for the ground unit depending on configuration. It’s expensive, but compared to a $200,000 light aircraft that needs a trained pilot, can’t fly from a rough bush clearing or function at night and is silent, it’s certainly a prospect.

As we discover over the next couple of days, the drone can easily stay up for an hour, and on one evening (when it wasn’t being confronted with 25 knot winds) for an hour and a half. It cruises at around 40 knots. You can easily programme search grids, point towards a point around which it should ‘loiter’, and quickly peel off from the plan and fly by wire.

I say fly, but Chris is careful to point out that you’re not really flying this thing. It’s flying itself, but when you’ve taken control you’re suggesting where it should fly. No matter how much you jammed the controls from corner to corner you couldn’t spin, invert or stall the Falcon thanks to its on-board autopilot.

We see lots of intriguing bright spots during our flight – the elephants, some probable impala, tiny pin-pricks of guinea fowl, as well as the more intended targets of a running junior warden. The second night trial we ran with the best infrared camera system that you are able to export from the US, a 480 x 640 pixel, 9hz system. Things were appreciably sharper, but with a slightly narrower field of view. As with any aerial surveillance, there’s a delicate balance to be struck between being close enough to recognise what you’re seeing and high enough to cover the area required.

The downside is the endurance. While range is not necessarily relevant – if your targets are much more than 5 km away through the bush at night your rangers are going to have a hard time getting there – it would be disappointing for the drone to have to return home before the ground forces arrived. But at least you’ve got a fix, and with a quick battery change you’d soon be back in the area.

As the battery charge began to run down – as indicated by a voltmeter on the laptop dashboard – Chris told the Falcon to come home with a click of the mouse. Soon the drone was circling overhead. Chis ordered it downwards and once he’d positioned it 50 metres above ground and slightly upwind he clicked another button and a parachute sprang from the Falcon’s rear and it drifted to the ground.

Drones go wrong, of course. Chris declined to give accurate statistics (though they must exist), but did say that small ones get lost more often than big, expensive ones. If a Falcon goes down, you’ve got its GPS that (you hope) was still recording and reporting its position, at least while it was above any trees.

The construction was rugged, but when Chris – flustered by media and on the last flight of a gruelling schedule of demonstrations – forgot to secure the camera payload properly on the last flight it detached when the parachute deployed and it went crashing to the ground. The expensive camera was safe but the carbon fibre housing cracked. There’s always room for human error.

Despite the hype around using drones for conservation they’ve yet to be used to any great extent. This may be about to change. Kruger National Park is now being targeted by poachers from across the border in Mozambique. There is currently no fence between Kruger and Limpopo in Mozambique, thanks largely to the substantial efforts of Peace Parks, who promote trans-boundary conservation.

Thanks to the poaching the Kruger authorities are talking about re-fencing the border and cutting the Kruger-Limpopo system into two again. In order to defend their concept, Peace Parks are purchasing a small fleet of drones that will be used in conjunction with ground and helicopter-deployed forces. They will be operated by UAV & Drone Solutions, a South African company formed by Otto Werdmüller Van Elgg. Otto was at the demo and we had long talks. He expects to be in the air and operational within three months. Others familiar with exporting drone-related technology from the US are less optimistic. Whichever way, this will be an interesting project to watch.

The US government is very wary of civilian use of drones, severely limiting their use within its borders and preventing the export of military-grade technology beyond them. Kenya is similarly cautious, and it’s hard to see how long it will be before they will be in use above East African landscapes. But from what I can tell, it’s only a matter of time.