Pairing drones with K-9 units on night pursuits across Australia
When a suspect flees into bushland outside a satellite suburb like Kellyville at dusk, the pursuit changes instantly. Streetlights disappear, the air cools, and officers must rely on what they can hear or smell. NSW Police and Victoria Police have long leaned on dog squads for this kind of work, but a thermal-equipped drone is now a regular partner, giving the handler a steady eye in the sky.
A trained malinois can follow a scent thread through long grass and across creeks, separating fresh tracks from residual odour hours later. What it cannot do is scan a two-hundred-metre radius for body heat or work for hours once adrenaline fades. At night the limits grow: dogs tire faster, scent behaves differently in cool, damp air, and twisted paws are common on rocky ground typical of regional Queensland and the Adelaide Hills.
Small unmanned aircraft fitted with thermal cameras help by sweeping a broad search area and feeding coordinates to the handler. The technology amplifies what the animal already does best, shortening the window in which a suspect can vanish into terrain that would otherwise swallow the trail.
Queensland's Moreton Bay pilots and Western Australian counterparts have been testing this pairing for several years during search and rescue, missing persons and post-burglary pursuits. Those lessons are now reshaping night operations training at state academies.
Why the combination works after dark
Drones offer broad coverage; canines offer confirmation. A thermal camera can pick out a person crouched beneath blackberries or pressed against a cold concrete tank, where torchlight would show nothing. The same camera guides the dog towards the most promising scrub, cutting search time from hours to fractions.
The value is high in Australia, where vehicle pursuits often span ten or fifteen kilometres before a driver bails out into state forest or farmland. Every minute saved reduces the chance of injury to the suspect, the officers, the dog or bystanders on neighbouring properties. In leafy Brisbane or Perth suburbs, where crossing a railway reserve is enough to vanish, the same logic applies.
Knowing roughly where the suspect is hiding also protects the dog. The handler can approach from the safest angle, position cover officers, and in many cases call for surrender before committing to an unguided advance through dense scrub. A controlled outcome preserves the dog's working life and lowers the chance of a bite that triggers state oversight review.
Sensor payloads suited to the mission
Most modern platforms used by Australian law enforcement, operated under a Remote Operator Certificate, already carry a stabilised thermal sensor. Radiometric thermal imaging lets operators read a target's temperature, distinguishing a living suspect from a recently extinguished engine or a sleeping animal.
Low-light starlight cameras extend hours into bright moonlight, often enough in regional New South Wales and Tasmania where artificial light is sparse. Where budget allows, laser rangefinders and zoom cameras let pilots confirm details without dropping below the altitude ceiling set by the Civil Aviation Safety Authority. Payload choice should track terrain: thermal dominates in the Otways' coastal forests, long-zoom visible-light cameras perform better across Riverina paddocks.
Operators should plan for redundancy too. Spare batteries and a swap point near the search sector keep the aerial unit aloft for as long as the dog squad needs it, especially in cooler southern states where cells discharge faster.
Australian regulation and operator responsibilities
Police drone work sits under the Civil Aviation Safety Authority framework, most often Part 102 for complex operations and Part 101 for many public-safety flights. Agencies typically hold a Remote Operator Certificate, with night operations requiring extra endorsements and night-rated pilots.
Beyond aviation law, agencies must consider the Australian Privacy Principles. Pursuit footage can record bystanders, neighbouring homes and private property, so retention periods, access controls and written justification for using recordings outside the immediate case are now standard. Publicly visible drone-use policies, similar to those already posted by several state forces, reassure the community that the technology is being used narrowly.
CASA may review an agency's documentation after a serious incident, and internal affairs routinely cross-reference drone logs with body-worn video and dog squad reports. Disciplined logging of launch times, flight paths and sensor modes is now standard hygiene, not an afterthought.
Training handlers and pilots to work as one
Joint exercises between dog squads and RPA teams are now routine in larger commands, often run with neighbouring agencies so cross-border operations do not depend on first-meeting relationships.
Integration centres on agreed communication: hand signals, radio phrases and abort cues that work when officers are focused on a dark treeline. Pilots learn to read the handler's calls and adjust altitude or thermal gain. Handlers learn to trust the pilot's readouts without losing their own situational awareness, an attitude built slowly through repetition.
Evening drills at facilities outside Nowra or Mount Macedon let both teams refine their rhythm until it feels instinctive. The investment pays off when a real pursuit crosses from a suburban street into adjacent bushland at midnight and the response has already been rehearsed many times.
Comparing day and night drone-supported tracking
The difference between day and night support is more than a sensor swap. It changes flight planning, team staffing and legal exposure. Reviewing both modes side by side helps commanders decide which deployments justify the cost of a night-rated pilot and which are better left to daylight shifts.
A blanket policy of always launching a drone rarely makes sense. Night missions need targeted justification, extra crew and slower decision cycles, matched against what would happen without aerial support.
| Factor | Daylight operations | Night-time operations |
|---|---|---|
| Primary sensor | Visible-light camera, often with zoom | Thermal or low-light sensor, sometimes paired with visible |
| Pilot visibility of airframe | Direct visual line of sight usually easy | Eyes must adapt to dark; night-rated required |
| Dog endurance | Higher, scent rises in warmed air | Lower, scent behaviour shifts in cooler, damper air |
| Terrain hazards | Generally visible from above | Hidden by shadow, especially rocky or flooded ground |
| Battery performance | Stable, slightly better in mild conditions | Discharges faster in cooler southern states |
| Risk to bystanders | Lower, clearer visual ID possible | Higher, identification harder without thermal cues |
| Regulatory layer | Standard Part 102/101 authorisation | Night endorsement and additional pilot currency required |
Practical steps for agencies considering this approach
- Establish joint drill programmes between dog squads and RPA teams before relying on the combination in live operations
- Choose sensor payloads around local terrain rather than buying the most expensive camera first
- Embed night-endorsed pilots within dog squad shifts so coverage is not lost when pursuits run late
- Document flight paths, sensor modes and retention decisions as standard after each deployment
- Publish an accessible drone-use policy that explains flight purposes, footage handling and complaint avenues
Agencies ready to formalise this work can begin by reviewing their current Remote Operator Certificate against likely night scenarios, then requesting a focused training week with a neighbouring jurisdiction that has already integrated canine and aerial teams. Start small, log everything, and grow from there.