Solutions

An intelligent and flexible system of advanced monitoring and inspection, using elements of artificial intelligence. The system is based on aerial drone fleets equipped with high-tech components, sensors and video cameras to collect operational information from monitored area.

Technically, the solution uses innovative algorithms that trigger software packages which process and re-render video images -similar to machine learning- for identification and automatic alarm triggering when incidents, defects, impediments, and intrusions occur in the field.

A computer-aided diagnostic system with a convolutional neural network algorithm that can fully detect prostate cancer using pre-biopsy magnetic resonance imaging. It is a non-invasive assessment of the prostate that reduces unnecessary biopsies and eliminates the inherent variation in performance and interpretation of MRI results.

An automated diagnostic system designed to perform cardiac imaging analysis, including calcium score quantification. Images obtained from CT and MRI are processed using cardiac segmentation technology, resulting in precise and immediate diagnosis, while significantly limiting need for feature engineering and expert knowledge.

System protection is carried out in the same way as current ICT implementations, but at a higher level of resilience, employing a Hybrid Intrusion Detection System (IDS) application, and an Assisted Moving Target management application for dynamic data network reconfiguration (port assignments, account configuration and file permissions).

Solution provides polymorphism protection with AI (Machine Learning Prediction & Anomaly Detection), polymorphism implementation and management application for function polymorphism APIs used by applications running on the system, procedural measures, and support for certified cryptographic equipment.

We use multisensorial 3D tracking to detect small unmanned air vehicles. Multi-sensor architecture exploits sensor modes including EO/IR cameras, acoustic arrays, and radars. These elements create an “acoustic camera” that performs spiral beam forming analyses using data fusion.

Using concurrent neural networks and a beam forming algorithm, the acoustic camera can detect and locate both stationary and moving noise acoustic sources. The resulting high–speed color map overlay and video are ideal for identifying transient sources of acoustic noise.

Mini-RPAS can depart from maritime platforms such as yachts. Their low cost and very small signature makes them a favorite platform for smugglers and terrorists. The mini-RPAS Radar Cross-Section (RCS) is too small to be detected by the regular costal radars, which is where SafeShore comes in. Another important maritime threat addressed is low altitude guided parachutes. Due to their low RCS, these parachutes are hard to detect by standard radars. They can be air-dropped away from the shore and can be used for deploying illegal items.

The SafeShore solution detects small targets that are flying at low altitude using 3D LIDAR that scans the sky and creates a virtual dome shield above the protected area. 3D LIDAR is integrated with passive acoustic sensors, passive radio detection and video analytics. With 3D LIDAR technology, SafeShore is able to cover a vast area of the coastal border using mobile platforms. Each mobile platform will cover its area with a dome-shaped virtual detection shield of a preset radius. Overlapping of platforms will allow for continuous detection shield along the shoreline.

Public video surveillance is becoming ubiquitously deployed as a valuable tool but denotes specific drawbacks, such as the absence of effective data processing, or the lack of advanced machine intelligence features like DROP-oriented (Distinctive Regions Of Interest or Patterns) retrieval architecture. Ready to replace traditionally generated analysis techniques such as face detection or color/motion information, DROP may include color regions, tattoos, logos or any other distinctive features that appear in given “incident” images.

Derived from real police video-based forensics investigations, where clear face identification of a suspect is not possible, Intellio could be used to identify and track relevant targets and target instances in the entire surveillance system using DROP retrieving algorithms. It is a dedicated solution optimized for running on low resources, embedded on hardware IP-based video cameras. The solution allows dynamic adjustment of algorithms running parameters to adapt them to different tasks in diverse environments.