When it comes to artificial intelligence (AI), science fiction writers have a field day writing apocalyptic stories of superintelligences that destroy the human race. But when one looks at the actual applications of AI in the real world, the picture is much rosier. Quite to the contrary of the doomsday visions that some imagine, AI-powered devices are making our lives better and safer.
Take the unmanned aerial vehicles developed in a partnership between Kongsberg Geospatial and SFL Scientific, for example. These drones are capable of autonomously scanning areas for a range of chemical, biological, radiological, nuclear, and explosive threats from the air.
While in flight, these devices have sensors to collect visual and thermal data. A FLIR MUVE C360 multichannel chemical sensor also provides the ability to detect various chemical and explosive hazards. Sensor data is analyzed via Generative Adversarial Networks and Graph Deep Learning models with a powerful NVIDIA Jetson embedded computer system.
Threat data, as interpreted by the machine learning models, is displayed in a ground-based control station, along with route suggestions made by the deep learning system. The route planning algorithms are designed to address any threats that the system detects. This information can assist first responders in identifying and prioritizing their responses to threat situations.
By offloading much of the analysis to the onboard AI systems, ground-based operators are better able to remain focused on their missions and be more effective.
Kongsberg Geospatial promises further details about their aerial vehicle in a free seminar on Wednesday, November 10th, 2021 at 3:00pm EDT. Register at no charge here.