The Nord Stream Pipeline sabotage that occurred in September 2022 highlighted how critical the underwater infrastructure is. The seabed infrastructure in fact traverses multiple countries and serves vital national interests.

Securing the infrastructure is not a piece of cake, because of its extension and location: cable-tracker TeleGeography estimates that in 2021 there were 807,000 miles of subsea cables in service globally, and the average seabed is in the range of 3.000 meters deep.

In this scenario the European Defence Agency (EDA) is evaluating proposals for a dedicated program for critical seabed infrastructure protection.

Unmanned underwater vehicles are a key asset for the infrastructure protection, and industrial players are targeting military and civilian customers for their unmanned underwater systems. But a new generation of UUV capable of reaching a deep of 6000 metres, corresponding to 97% of seabed, is not enough to solve this challenge

A critical factor is represented by research in new sensors and automatic target recognition to classify potential threats as early as possible. Artificial Intelligence deployed at the edge in a distributed IOT pattern (whether in a UUV or static sensor frame) is a promising solution to this challenge.

Within this scope the EU program InSecTT aims at bringing Internet of Things and Artificial Intelligence together.

Part of ISecTT is a smart underwater surveillance solution for seaport waterside protection to provide a timely and effective response against the surface and underwater potential threats.

But bringing AI to the edge in demanding environments (UUV or underwater sensors) is a challenge by itself.

At E4 we have partnered with the company Aitech Systems.

Aitech Systems produce una serie di computer rugged per inferenza di reti neurali, progettati per funzionare negli ambienti più estremi.

Aitech Systems produces a set of rugged AI inference computers designed to work in the most demanding environments. The Aitech Systems A179 inference compter is based on Nvidia Jetson NX, and with a weight of just 600gr. is ideal for automatic threat classification on a remote underwater sensor or UUV.

The Aitech Systems A178 inference computer is based on Nvidia Jetson AGX and brings tremendous detection and classification performance at the edge.

E4 Computer Engineering can support the whole workflow of an Artificial Intelligence application.

With GAIA, the E4 platform for enabling scalable and high performance Machine Learning and Deep learning workloads, and with flexible small form factor inference computers for industrial or rugged applications.

For further information, please contact us