h-aero above a forest landscape

High-tech for better forest management

More data about your forest — captured from the air for hours, right above the canopy, low-noise and without rotor downwash.

The problem

Forests under pressure

Drought, bark beetles, windthrow and a rising fire risk are taking their toll on our forests. Conventional surveying reaches its limits: ground surveys are point-based, satellite data is coarse, and battery-powered multicopters spend most of their energy on hovering, which limits their time over the area. Foresters need current, area-wide and reliable data as a basis for their decisions.

The solution

Capture → Analyze → Decide

A clear workflow — from the low-noise flight above the canopy to a reliable basis for decisions.

Step 1

Capture

Low-noise flight directly above the canopy — high-resolution, georeferenced and with long endurance: more than 5 hours, measured with a 1 kg payload in economy operation.

Step 2

Analyze

AI-assisted analysis of the aerial data: damage areas, vitality and structures become visible.

Step 3

Decide

Reliable maps and reports as a basis for forest planning and measure planning.

Why h-aero

The energy belongs to the payload

A multicopter spends most of its battery energy on hovering. With h-aero the helium carries the system and the energy is available to the sensor. More than 5 hours have been measured, with a light payload in economy operation.

Quiet and without downwash

No rotor downwash stirring up crowns, soil or sampled air, and a noise level below 40 dB — important in protected areas, for wildlife observation and for measurements close to the ground.

Experienced over forests

h-aero holds flight-ascent permits for operations over forest land and is active in the EU project FORWARDS (Horizon Europe) with its HALO-FARM sub-project (see the data chain).

Funded by
the European Union
HALO-FARM has received funding from the European Union's Horizon Europe research and innovation programme through the FORWARDS project (Grant Agreement No 101084481), EFI grant G-07-2024-3. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Executive Agency (REA). Neither the European Union nor the granting authority can be held responsible for them.

HALO-FARM · EU project FORWARDS

From satellite view to close-up: the data chain at site 1

HALO-FARM surveys five forest sites in Germany and Austria as free-balloon strips of 1 km × 10 km (5,000 ha, about 3,900 ha of forest). The satellite finds changes over large areas; h-aero delivers the close-up for expert interpretation. Six stations, shown here for site 1 near Baden-Baden — what already exists is marked real, what is scheduled is marked planned.

  1. 1 Selectionsites and strips
  2. 2 Satellitetime series 2025–2026
  3. 3 Samplesown imagery
  4. 4 Simulationflight plan
  5. 5 Free driftalong the strip
  6. 6 Analysisthree examples

1 · Selection: five sites, five strips real · Oct 2026

Demonstration — your changes are not saved. Turn a strip to see how its forest share changes.

Five sites: 1–3 near Baden-Baden, 4 near Dornbirn (Austria), 5 in the Rhön. Each site has a centre and a release direction chosen for the largest forest share. Tethered samples are taken inside the 1 km circle; for the survey the free balloon is released at the centre and drifts with the wind at constant altitude along a 1 km × 10 km strip. Flight altitude = strip maximum + 100 m (792 / 808 / 975 / 1,254 / 985 m a.s.l.). Site 1: strip 135° (SE), about 90 % forest, terrain from 270 to 683 m. Together the five strips cover 5,000 ha, about 3,900 ha of it forest (OpenStreetMap).

HALO-FARM KML Selector in free-balloon mode: release circle at site 1, strip towards the west with four waypoints and their altitudes
The HALO-FARM KML Selector — the entry gate to every location in the digital HALO-FARM world. On the flight day the strip follows the forecast wind: the Selector sets release point, direction (here 267° W for 1 November 2026), waypoints and altitudes and checks them against the terrain — the plan that station 4 simulates. Demo on request.

2 · Satellite view: monthly time series since April 2025 real · Apr 2025 – Oct 2026

Sentinel-2 (Copernicus), one low-cloud scene per month over a 3 × 3 km window: 17 of 19 months usable. Right half: vegetation index (NDVI) of the forest; magenta marks forest whose NDVI dropped by at least 0.15 against the same month of the previous year (May–September only). The largest value — 1.4 ha inside the circle in August 2026 — is a candidate for a close-up, not a diagnosis: that is what stations 3–6 are for.

Time-lapse April 2025 – October 2026 with project start (3 June 2025) and our own imagery picture-in-picture
Film strip of the monthly Sentinel-2 scenes of site 1 with project months
All usable months at a glance, with project month (M1 = June 2025)

Prepared with our Open Hub satellite viewer data chain (viewer access on request). Data: Copernicus Sentinel-2 (ESA).

3 · Samples: our own imagery at the sites real

4 · Simulation: plan the flight before the flight real · since Oct 2026

Our flight simulator checks each strip against terrain and wind forecast and optimises the free lift and the valve times so that the planned altitude profile is followed. Example for site 1 (planned flight on 1 November 2026, forecast wind): 7.7 of 7.9 km in 27 minutes, on average about 19 m from the planned altitude profile, 24 g of helium. The lateral offset of up to 1.4 km shows the wind turning with altitude — exactly what this step has to reveal before a campaign.

Simulator, free-balloon mode, site 1: the whole 27-minute drift in 22 seconds — track over the strip above, planned and simulated altitude below (simulation, not a flight). Demo on request.

5 · Free drift along the strip planned · first forest flight early November 2026

Free-balloon operation of the same platform was demonstrated in 2026 outside HALO-FARM. Over forest, the strips are flown from early November 2026 at site 1, followed by leaf-off (winter) and leaf-on (spring/summer) campaigns at all sites in 2027.

6 · Analysis: three examples from our own imagery

Detection methods are being developed on our own data. The examples show the direction of the analysis work, each with its origin.

Thermal image of the meadow at the forest edge, top original, bottom with the candle circled and zoomed
real · M3 Smallest heat source. A candle on the meadow at the forest edge of site 1, found automatically in 86 % of the frames of the thermal livestream (31 Aug 2025) — the analogue for early wildfire detection.
Aerial view in dense fog, top original, bottom contrast-enhanced with a group of people marked
real · M9 People in a foggy winter forest. Demonstration flight at site 2 (25 Feb 2026): contrast enhancement makes a group by the road visible that the fog hides (local contrast × 2.9). People are not identifiable.
Drone view over site 1 in late summer 2020, top original, bottom with brown crowns marked in magenta
before project · method Brown, dying crowns. Our own drone video over site 1, late summer of the drought year 2020 (Battert ridge in the background). A simple colour threshold marks the brown crowns: 7.3 % of the forest seen in this frame (uncalibrated share of the image) — the first step towards the crown-vitality analysis.

Status October 2026 (project month M17). Dates are given as recorded; material from before 3 June 2025 is marked as preparatory work.

Technology & sensors

Three sizes, one platform

The payload capacity is the shared budget for sensor and battery. The heaviest battery that still fits the budget next to the sensor gives the longest time over the area. We are happy to run the numbers for your scenario.

h-aero® zeroh-aero® zero+h-aero® one
Payload capacity (sensor and battery)500 g1,000 g3,000 g
Typical sensors4K camerathermal cameraseveral sensors
Flight endurancemore than 2 h (design)more than 5 h, measured with a 1 kg payloadmore than 5 h (design)

Flight endurance depends on the sensor's energy demand. The figures are maximum measured or design values with an economical payload, not typical values.

Three modes of operation

On the tether

As a tethered balloon over an area, in Germany permit-free up to 30 m, above that with a flight-ascent permit. For keeping watch over an area, for example after a fire.

Guided with propulsion

With its own propulsion over small areas, at low height and within sight of the platform. For samples and close-ups at a stand.

Free drift

As an unmanned free balloon, covered by our flight-ascent permits, along a planned strip: the platform glides with the wind across the area and produces a continuous picture of the situation.

Like any balloon, the platform needs a weather window. In return it then stays over the area for hours. Which mode fits depends on the area, the question and the permit.

Sensors the platform has already carried

The platform is sensor-agnostic. These payloads have already been on board, some of them as our partners' instruments.

Interfaces: USB-C · LAN · Wi-Fi (5 GHz) · mobile network (LTE, 5G-capable) · SatCom. Noise level below 40 dB.

Applications

From the forest-management spectrum

One platform, many fields of use — wherever area-wide, gentle capture is required.

Focus: wildfire

Where battery drones have to land after minutes, h-aero stays in the air: as a fire watch in high-risk conditions and, after the fire, as a watch over the area, on the tether and low-noise. The thermal camera on board reveals hotspots that cannot be seen from the ground.

Bark beetle & pest infestation

Map infested areas (red- and grey-crowned trees). Early detection of green-attack stages is developed with partners in a pilot.

Vitality & forest-condition monitoring

Seasonal and large-area capture of canopy condition.

Storm & windthrow damage

Map damage areas and canopy gaps after storm events.

Early wildfire detection

Thermal hotspot detection and persistent fire watch in high-risk areas.

Post-fire hotspot monitoring

Monitoring of burnt areas after extinguishing, until smouldering nests are out, with thermal imaging from the air.

Forest inventories & stand-structure mapping

Carbon inventory and taxation parameters of the forest inventory — captured from the air with 3D/LiDAR support.

Reforestation monitoring

Documentation and monitoring of calamity areas after afforestation and reforestation.

Wildlife monitoring

Inventory of the game population and monitoring of wildlife movement — low-noise and without startling wildlife.

Geostationary base station

Stationary use as an airborne base station — to improve positioning and radio coverage over the forest.

An added advantage in sensitive protected areas and national parks: low-noise, downwash-free operation that does not startle wildlife.

Deliverables

What we deliver today. And what we develop with you in a pilot.

Foresters have worked with remote sensing for decades and know its limits. That is why we draw a clear line between what we deliver today with evidence and what emerges in a pilot project with specialist partners. All analyses follow forestry definitions and the data of the forest inventory.

What we deliver today

Proven in our own operations

  • Georeferenced aerial imagery (RGB, thermal) and video as raw data for your own analysis
  • Live picture from the air with thermal imaging and hotspots, visible on the ground during the flight
  • Orthomosaic of small areas as a stitched map, with a note on stitching artefacts over moving crowns
  • Mapping of visibly damaged trees red- and grey-crowned trees, windthrow and gaps from the aerial imagery
  • Satellite time series of your area monthly since 2025 as lead-in and context for the flight

What we develop with you in a pilot

With specialist partners, checked against reference measurements on site

  • Tree heights and crown models by forestry definition on the upper stand, not as the highest pixel
  • Tree species, mixture shares and stem count
  • Timber and CO₂ stock
  • Early detection of bark-beetle infestation on still-green trees
  • Wildlife counts by thermal imaging in the leafless winter
  • Automatic detection of fire hotspots
  • Acoustic monitoring microphones have been carried, the analysis is open

The benchmark for every pilot product: accuracy against the reference data, area output and cost compared with the survey on foot. We define the scenario together before the first flight.

Process

From first look to monitoring agreement

  1. Initial assessment — a joint evaluation of your areas and your monitoring needs.
  2. Pilot flight — low-noise capture above the canopy and analysis of the aerial data.
  3. Monitoring/service agreement — recurring, reliable capture as needed.

Next step

Experience a demo or plan a pilot flight?

Get to know the platform at a free demo in Baden-Baden, or talk to us about a pilot flight over your forest areas.