The World - Represented by data

All action generates data, directly or indirectly through its impact on other systems.

This data represents the behaviour and status of people, machines and systems. Data has the potential to enable us to learn what the world looks like when things are normal and then immediately detect when anomalies occur, to detect impending disaster or predict system failure. This real-time insight enables action to be taken at the earliest possible moment. 

 

The volume and velocity of this data is set to increase dramatically in the near future as systems like autonomous vehicles and the Internet of Things become mainstream. This creates an opportunity to understand the world even better, but also greatly increases the challenge to make sense of data that may seem to be chaotic.

Rich intelligence is there, ready to be discovered

Persistently monitoring and analysing information becomes increasingly difficult for humans as the duration, diversity and volume of data increases. Humans find it hard to maintain concentration for extended periods of time, to provide continuity between work shifts, or to process big data streams and  large varieties of data. 

 

Cyberupt solutions are designed to fulfill this role and allow the human to focus on that which is most interesting, most threatening or on that which specifically need human insight. Our systems augment human judgement by avoiding confirmation bias and potential under- or overreaction to events; attributes which are highly important for operations in complex environments.  

Cyberupt solutions

Multi-source intelligence creation

Cyberupt solutions fuse multi-source enriched data to create a single view of the observed physical or virtual domain. This enables persistent, real-time insight over extended periods of time into behaviour, activities or transactions based on a broad range of inputs, thus enabling immediate reaction. Using statistical modelling new observations are tested against an expectation, allowing the discovery of patterns, anomalies, trends and developing threats. The system automatically learns from observations to “understand” what is normal using Model Based Machine Learning, so as to detect that which is unexpected or indicative of a threat, even in a potentially fluid environment. Contrary to learning methods used by other Artificial Intelligence approaches, Model Based Machine Learning is well suited to the detection of rare events and anomalies.

Distributed Datacentre

Flexible, Scalable, Robust, Distributable

Information to the Right people, at the Right time, in the Right format

Data ingested from any connected sensor and information source, is enriched and probabilistically fused to provide a comprehensive, mutidimensional view on the entity, transaction or event being observed. Human input (HUMINT) is incorporated where available to further enrich information.
 
Statistical modelling and Model Based Machine Learning services use this fused data to discover intelligence, which is then in turn made available to the user at the edge. The System comprises of an Infostructure which uses interconnected Data Centres as building blocks. 
The system uses a Service Oriented Architecture (SOA) which enables seamless extension with new services and information sources and to scale from small to very large deployments. The Data Centres are connected using robust, encrypted links and data is transparently made available across the entire network, enabling users to connect to the most convenient node while accessing any information they need. The information is presented on the desired user device or geospatial platform, which may be in use by a dismounted person, mounted in connected vehicles or aircraft or in operation control centres.

Solutions in many domains

Whether it is about protecting a maritime environment from illegal fishing, illegal bunkering, piracy or environmental damage, or about discovering financial transactions related to criminal activity or about understanding the network of collaborators involved in terrorist activities, the ability to discover patterns or unusual behaviour amongst billions of events enables the end user to act pre-emptively and with knowledge.

Cyberupt BV

Cyberupt BV is based in The Netherlands and provides solutions for UN recognised government agencies and organisations around the world.

 

Our engineers and data scientists have an extensive track record in the development of own technology and products. Our skills include the mining of massive data sets, machine learning, probabilistic modelling and the design, development and deployment of very large distributed systems.

 

Cyberupt has an operating philosophy of partnering with its customers and key suppliers to maximise operational benefits of its systems.