We incorporate Nvidia's most powerful AI computation module in our IA4Birds project

By The AIR Institute

The Jetson AGX ORIN, Nvidia's most powerful model, is the AI computing module chosen for the electrical panel that will be part of the bird detection tool designed for the IA4Birds project, coordinated by AIR Institute and funded by Fundación Biodiversidad. The aim is to use artificial intelligence to use audiovisual devices to monitor bird populations so that, based on a more in-depth knowledge of their behavior, it can prevent threats and decide whether or not a site is suitable for the installation of a wind farm.

This is the high-end of the AI computing modules; a supercomputer at a reduced size and with better energy efficiency than the previous model.

Ismael García and Guillermo Redondo, researchers at AIR Institute, have been in charge of the 3D and electronic design and component selection for the electronic panel of which the Jetson AGX ORIN is part, together with an Uninterruptible Power Supply System, an air conditioning system and the casing to protect the equipment, since it has to be in the open air.

Specifically, the electrical panel includes:

  • Jetson AGX ORIN 32GB: NVIDIA AI computing module that provides AI processing capabilities at the edge.
  • UPS (Uninterruptible Power Supply): This device is crucial to ensure that the system continues to run in the event of power outages. The UPS is housed in such a way that it will provide a backup connection to all critical components, with batteries that will remain charged and ready to take over in the event of an interruption to the main power supply.
  • Climate control system: To enhance outdoor operation, the climate control system will maintain optimum internal temperature and humidity inside the enclosure. This is achieved by means of fans, heat exchangers and heating elements. The internal organization of the electrical enclosure allows for adequate airflow for climate control and sufficient space for the necessary wiring and connectivity for all devices.
  • Outdoor-rated enclosure: The enclosure itself is designed to be resistant to outdoor conditions. This involves robust construction with corrosion-resistant materials, seals to protect the equipment from water and dust (with an IP rating suitable for outdoor use), and a design that allows for easy installation and maintenance access.

Attention has also been paid to the functional layout of components to facilitate inspections and maintenance. In this regard, the model shows clear wiring routes and modular components, which suggests that the design will allow for easy expansion or modification depending on the future needs of the IA4Birds project.