Milestone Systems and Aalborg University join technologies to save lives

The project between this academic entity in Denmark and Milestone Systems promotes the AI capabilities of its thermal cameras and video technology software to monitor behavior and save lives in port areas.

The Aalborg University, in Denmark, has been recognized for its advances in port security technology through a collaborative project with the video technology provider Milestone Systems, in which the research teams have developed a model that can train itself to monitor and assess abnormal behaviors and prevent accidents, For example, near the water.

In Denmark alone, urban ports have witnessed numerous drowning incidents over the years. Este proyecto sobre la detección accidentes mortales en puertos utiliza modelos entrenados con inteligencia artificial (IA) para cámaras térmicas.

La tecnología de vídeo y los conjuntos de datos sintéticos implementados en un puerto danés han recibido el Premio de Investigación de la Fundación Spar Nord, por “promover las capacidades de inteligencia artificial de las cámaras térmicas y el software de tecnología de video de Milestone para monitorizar y evaluar comportamientos inusuales y mejorar la seguridad en los puertos de todo el mundo”, según el comunicado conjunto.

Tras identificar una clara necesidad de prevenir este tipo de accidentes, A research team from Aalborg University employed AI combined with video technology to improve safety in one of Denmark's busiest ports,, Aalborg's port,. Thus, they created he the largest outdoor thermal dataset for video analysis, spanning a period of nine months.

The action of falling into water is a specific activity that cannot be easily reproduced in the real world, since volunteers cannot be asked to jump into water for safety reasons.

Using a combination of a crash test dummy filled with warm water detectable by thermal cameras and the largest published collection of annotated thermal images, se entrenaron modelos de IA para detectar y alertar rápidamente a los equipos de rescate si una persona caía al puerto.

El equipo de investigación creó un conjunto de datos sintéticos avanzados que podrían entrenar eficazmente a los modelos de IA en los movimientos repentinos e involuntarios que implican las caídas. Los datos de entrenamiento se ampliaron para incluir usuarios de sillas de ruedas, patinadores y ciclistas, con el fin de proporcionar a este modelo inteligente tantos escenarios diferentes como fuera posible.

How do you explain Neelu Madan, investigadora principal y estudiante de doctorado en la Universidad de Aalborg, “al conocer el comportamiento normal en la zona ribereña, el modelo puede detectar eventos anormales, como una persona que se cae del muelle, and immediately trigger an alarm. This way, rescue personnel can be called quickly, even in cases where the accident occurs without witnesses.

In addition to helping save lives, this award-winning research demonstrates the power of synthetic data used to train AI models in situations that are difficult or unsafe to replicate in the real world.

'We are incredibly proud of Neelu Madan and excited about the potential application of her work,' says Rahul Yadav, Chief Technology Officer of Milestone Systems. The development of the thermal dataset at the port of Aalborg represents a revolutionary breakthrough in video technology and security.