LoRaWAN IoT & GeoAI

An overview where mobile coverage cannot reach:
Capture location and condition – on alpine pastures, in forests, across open land.

VirtualShepherd combines LoRaWAN sensors with spatial AI analysis: reliable location and condition data from areas where conventional radio links do not carry.
Expansive areas, steep terrain, dense vegetation — mobile coverage is unreliable or absent in all of them. VirtualShepherd therefore relies on LoRaWAN: low-power, long-range and robust against topography. Sensors report location and condition across kilometres to a gateway, the data flows into the GeoAI analysis from MovingLayers and becomes maps, movement patterns and alerts. Instead of inspection rounds and guesswork, an up-to-date operational picture is available — for animals just as much as for machinery and installations.
LoRaWAN radio network Without mobile coverage GeoAI analysis Virtual fences Alerts in real time
70%
less effort spent looking for animals
Learnings from the grazing season
2 min
to find an animal on the map
Feedback from an alpine farmer
1 glance
across everything happening on the land
Alpine pasture management
Subscription
across the alpine season – pay only while you use it
Low-risk entry

VirtualShepherd App

The app shows every sensor on the map: location, movement and alerts — even when you are not on site.

  • Location data over LoRaWAN
  • Virtual fences with notification
  • Density maps and movement patterns
  • Operable with gloves in the field
Book a free demo now

Geodata from field to decision – standardised, versioned & critical-infrastructure-ready.
Consistent layers & symbology · Full version history · Operated locally & auditable

Radio that arrives where mobile coverage gives up

On alpine pasture, in forests and across wide areas, mobile coverage often ends exactly where the work begins. VirtualShepherd therefore uses LoRaWAN: a narrowband radio network that carries very far on very little energy and penetrates terrain, vegetation and hollows far better than mobile networks. A single gateway can serve an entire valley. How long a sensor lasts without recharging is determined by the transmission interval and the volume of data – at the reporting rate common in alpine pasture management, one charge lasts across the full season.

Location and condition can therefore be captured even where, until now, only walking the ground gave certainty.

Radio signals become decisions

Raw position reports help nobody. VirtualShepherd brings the sensor data together with your geodata and analyses it spatially: where does the herd prefer to stay, which areas are avoided, which movement falls outside the pattern? Ongoing measurements become density maps, movement histories and alerts — and with them the opportunity to act before something has happened.

Virtual fences add to this: you define areas on the map and are notified when an animal leaves one. No fence building, no inspection round.

What the system does today

  • Create animals and equipment and pair them with sensors
  • Precise position — including in remote areas without mobile coverage
  • Density map for preferred and avoided grazing areas
  • Virtual fences with notification when an area is left
  • Real-time alerts – on the smartphone and on the web
  • Transmission interval and data volume configurable per use case – they determine how long the sensors last

The solution grows with practice: new functions emerge in close coordination with the alpine pasture community — driven by application rather than drawn up at a desk.

View use cases

Not only animals

What works for alpine livestock carries anywhere that location or condition matters and no reliable mobile connection exists: machinery, equipment, installations. The same radio network, the same analysis, different objects.

The data is used today in alpine pasture management, in agriculture and forestry, in tourism and local recreation, in research and by insurers.

Where VirtualShepherd comes from

The system emerged from an FFG-funded research project together with the University of Salzburg. What began there as a question about reliable positioning in mountain terrain is now in practical use on the Zeiselalm in Lower Austria and the Lorenzalm in Carinthia, developed further along what alpine farmers actually need.

Where VirtualShepherd sits in the geodata lifecycle

VirtualShepherd is one of the two forms of the capture phase: the data stream that runs without a crew and without a planned end. The second form is the site visit. Often it is the data stream that first shows a site visit is needed.

  • 1 · Capture by site visit: MoversSurvey MDE – structured field data capture for iOS, Android and the web.
  • 2 · Validate and prepare: Operations and maintenance – validated before it counts.
  • 3 · Understand and decide: Kermit GeoAI – answering spatial questions in plain language.
  • 4 · Implement and improve: the decision becomes a measure – and because the measure changes reality, it writes the capture brief for the next cycle.

That is how the loop closes and the next cycle begins. View the geodata lifecycle

Less searching. More overview. Reliable data from difficult terrain.

MovingLayers combines LoRaWAN sensors with spatial analysis — so you know what is happening on your land without walking every slope yourself.