From Teamwork to Dreamworks
- Dave Black
- Mar 31
- 5 min read
We refer to a ‘bird’s-eye-view’, but the latest technology is increasingly making a 'bee’s-eye-view' better understood to researchers, thus improving knowledge of the movements of our honey bees. Dave Black views the videos – and you can too! – to assess this technology’s value to beekeepers, scientists and even hornet-hunters.
By Dave Black
The ink had hardly dried on the page for last month’s exploration about drones and tracking technology when I found a pre-publication article about the latest equipment. The paper, actually published on March 9,[i] combines some fancy, advanced optical tracking with drone robotics to observe wild insect behaviour over several kilometres. The system, called Fast Lock-On (FLO) tracking was first proposed in 2024 by a team at Freiburg’s University in Germany,[ii] a development of six year’s work.
Tracking small things can be ‘active’, in which case they have to be able to carry a portable (battery-containing) transmitter, or ‘passive’, in which case they must be able to reflect light, (or another electromagnetic transmission) back for an observer to see it. Each method has advantages and disadvantages, but the important feature is range. A small insect can’t carry much, so the range is usually limited, but in theory if there is a device that can follow the insect closely you could actually track the device and not the insect itself, and your device can carry a heavier transmitter.
Tracking, after all, is really just ‘watching’. When things are small and moving, the resolution of your eye or camera starts to matter, and you have to be able to focus and move (the camera) continuously. Computers can do that quite quickly, better than people, but it’s necessary to simplify the image as much as possible to allow them to manage the movement bit without being overwhelmed by calculating with data from up to 1000 images per second.
FLO – From Golf to insects
FLO tracking uses a camera to determine the location of a specific object and then follows it. If you watch golf or nature programmes you will know how well these systems can work. To use the system for following something as small as an insect you use a bright paint marker, or better, a ‘retroreflector’ (like the ‘cat’s-eye’ in the middle of the road) so the subject is easy to pick out from the background which, for tracking purposes, is irrelevant. FLO tracking monitors the position of the object it has ‘locked-on’ and steers the camera to keep it ‘locked’ to the image. So, when the camera is on a drone (a quadcopter UAV), the drone will follow the bright image produced by the marked insect. Drones can fly a long way and are easily tracked; it’s routine for a drone to ‘know’ its position using GPS satellites.
The Freiburg team produced several videos including one following a honey bee flying from a release point in a field back to its hive. Because there can be more than one camera ‘on board’, the system display can show different live views simultaneously; one of a position on a map, a full colour view of the landscape covered, and a ‘bee’s-eye’ point-of-view. If you follow the references list below, or at this link, you can check out the videos yourself.
The radio tagging equipment available as used in Auckland costs about €2400, plus taxes and shipping (about NZD4,700), for a kit including transmitter (5), receivers (3), charger, power supply, a scale, and various tools, jars and so on. So, not particularly expensive. A reasonable estimate suggests that the tags have a range of about 250m-500m, and that can be increased by using a drone relay to about 800m. Surprisingly, the FLO tracking equipment costs are not dissimilar, but it isn’t a handy pre-prepared, kit-set-in-a-box – yet. Its imaging is a great improvement on harmonic radar, the data is three-dimensional, and the drone will steer around anything blocking its view and stay on-target. The system (accidentally) detected two flights at least 900m long, but in this study tracking distance was not the point.
A Tool for Hornet Hunting?
The range of both, for the purpose of hornet tracking, is constrained by something more mundane, the world of property rights. Racing along behind, hurdling fences and scaling buildings at the pace of a world-class marathon runner isn’t on the cards for hornet-chasers, and that’s why the radio kit supplies three receivers. The tracking aces can follow and use a vehicle, handing off to a colleague with another receiver before it goes out of range. Nor can I imagine watching a UAV following a hornet through Auckland’s back-yards and streets five or six metres off the ground. Despite what some scientists think, UAVs do not enjoy ‘unrestricted mobility’.
Plotting Unique, but Precise Paths
In natural landscapes or large agricultural spaces the new system looks very promising. In the paper just published the Freiburg scientists used the system to follow 255 flights to and back from a feeder positioned 122m away from a hive in a maize field. They were interested in the kind of navigational strategies the bees would use, landmarks, the path(s) they had travelled, a mental ‘map’, or vectors relative to something like the sun. They found the routes the twenty-six bees travelled were all very individual. Each bee used visual information differently, and will have selected a different strategy to suit various parts of the trip.
Generally, the bees seemed to be using a vector (an angle) outbound to the feeder, so the specific route could meander because the destination was invariable. They could take different paths around a tree, for example. Each bee had its route, but it wasn’t always the same one. Inbound flights (homeward) were much more consistent, straighter, nearly always the route the bee had used last time, and a lower altitude. Any route was more consistent when there was a landmark. Part of the reason homeward flights were more consistent maybe because the bees were very familiar with ‘landmarks’ at the hive. Going home, even if they were using a memory of their outbound path, they could at least correct it because they could see the right landmarks.
The flight data revealed very precise navigation, far more precise than honey bee’s notoriously error-prone waggle dancing might suggest. However, their individuality, and the range of possible landscape permutations, suggest that even with highly capable tracking equipment it will be a very difficult thing to investigate well. In this particular report the landscape was, well, pretty local and uncomplicated, and that is probably reflected in the behaviour they found. It might still be great to develop its use on those pesky hornets though. Tracking could be as fun as a video game!
Dave Black is a commercial-beekeeper-turned-hobbyist, now retired. He is a regular science writer providing commentary on “what the books don't tell you”, via his Substack Beyond Bee Books, to which you can subscribe here.
References
[i]Stentiford, R., Harrap, M.J.M., Titov, V.V., Lochner, S., Straw, A.D., 2026. Precise, individualized foraging flights in honeybees revealed by multicopter drone-based tracking. Current Biology S0960982226000849.
[ii]High Resolution Outdoor Videography of Insects Using Fast Lock-On Tracking, 2024. T. Thang Vo-Doan, Victor V. Titov, Michael J.M. Harrap, Stephan Lochner, Andrew D. Straw. Science Robotics 9. https://doi.org/10.1126/scirobotics.adm7689
Comments