Sensor-based AI, energy-efficient electronics and satellite-based IoT technology for wildlife research and conservation

Our motivation

Our world is changing - advancing climate change is destabilizing our ecological balance, global trade, increasing world population and the clearing of forests purely for the purpose of agriculture and livestock are leading to unforeseeable consequences for humans and animals:

  • More and more animals are being displaced from their ancestral habitat in their search for food and are invading human habitat.
  • Introduced animal diseases and epidemics threaten not only humans and animals in their country of origin, but increasingly also our immediate environment.

To protect biodiversity, the Fraunhofer Institute for Integrated Circuits IIS and the Leibniz Institute for Zoo and Wildlife Research (IZW) have joined forces in the GAIA-Sat-IoT Project.

With the help of technological innovations and scientific expertise, we are generating new synergetic solutions. Our goal is to develop camera tags equipped with sensor-based AI and satellite-based IoT communication, intended for bird and wildlife tagging.

This will allow us to detect, document and counteract animal behavior to ecological changes and processes at an early stage.


The challenge

Hardware and energy-autonomous operation
The camera tags must be adapted to the respective animal species and must not obstruct or interfere with the animals in any way. The use of the animal transmitters extends over a longer period of time and over long distances. This results in limitations in terms of weight, size and energy budget.

Data aggregation
Artificial intelligence can be found in more and more areas of life and applications, but training neural networks requires both large amounts of data and high computing power, which comes with the corresponding hardware. The datasets required for the field of wildlife research are not sufficiently available and arbitrarily reproducible. The final trained algorithms are usually very computationally intensive, making this difficult to execute on small, portable embedded systems where little computing power is available.

Data processing and transmission
The processing of sensor and image data and the data transmission via satellite require a lot of energy and storage capacity and cannot be implemented on the necessary small computing platforms.



This is how we proceed

Processing of the data on the (animal) transmitter
Since the transmission of raw data via current satellite communication networks is too energy intensive, we are working on image processing close to the sensor. The key advantage of processing the data directly in the computing unit is that decisions can be made in real time about which data to send to the satellite. The utilization of the radio channels is to be optimized by the further development of artificial intelligence to the extent that only the image and sensor data that provide relevant information are transmitted. The improved image recognition AI will assist in identifying animal behavior patterns relevant to the particular use case and triggering/triggering the camera sensor only when needed.

Transmitter development
As a first step, we are developing a miniaturized animal transmitter that has a camera module, various sensors, and a local memory unit. Since the transmitter will be used early in the project to collect training data, radio transmission is (still) limited. In the next step, we will further develop the transmitter and integrate the AI signal processing electronics. Another essential part is the integration of a powerful satellite IoT radio module into the transmitter to guarantee the later transmission of the extracted information. As a result, a small animal transmitter is created, which can intelligently pre-process and combine data and send all relevant environmental information through a communication link.

IoT communication structure
Especially in areas beyond the reach of terrestrial communication infrastructures, the support of satellite-based networks is required. To enable direct transmission from the transmitter node to the satellite, we are developing a communication system based on a terrestrial mioty® technology. We adapt this method to the satellite scenario and thus ensure robust and energy-efficient data transmission. In addition to the transmitter development, the receiving side, in this case the satellite, must also be adapted in order to be able to receive and process messages there.



Our contribution

  • Fraunhofer IIS - Division Smart Sensing and Electronics
    • Embedded software (microcontroller software), e.g. intelligent image acquisitio
    • Miniaturized, energy-efficient sensor technology and hardware, e.g. time-controlled camera triggering, adapted design
    • Sparse AI, e.g. energy-efficient algorithms
  • Fraunhofer IIS - Communication Systems Division
    • Embedded software: development and design of the radio module on the transmitter and its counterpart on the satellite
    • Mature IoT technology mioty® from terrestrial deployment
      • Frequency adaptation of terrestrial frequencies
      • Direct transmission of data via satellite
      • Receiver architecture for distributed LEO satellite systems
      • Doppler compensation on the satellite