Intelligent Observation Systems

Challenges

 

In any area of business, there are experts with great knowledge, who are crucial in making complex decisions. These experts assess the environment to provide the best response according to the ideal strategy. Many times, key processes within a company rely on these types of experts, whose knowledge is difficult to convey.

The main challenge in these cases is to capture high value knowledge and extend it to the entire organization.

 


Approach

 

AIA’s approach goes through the automation of decision-making processes. The automation of decision-making by experts requires Artificial Intelligence techniques to:

  • Capture expert knowledge
  • Emulate human reasoning to automate decision-making

AIA has technologies and methods of expert knowledge extraction and emulation of human reasoning. An SIO is an evolved artificial intelligence technology that emulates the way to look of a living. The cognitive process of the living world is divided into two parts: the outer and inner world. The outside world only captures a portion of the information through specialized sensors. At this time, there is no knowledge, but only received information.
Knowledge or understanding only occurs when we have been able to provide meaning to the information received. That is, when we are able to build abstract concepts from the data. These abstract concepts are the way to knowledge and are part of the inner world.

 

 

 

 

 

 

To construct an SIO it is necessary to define a particular conceptual structure associated with data flux. The SIO will be able to analyse this data in the same way as the expert who designed it.


Benefits

 

SIO provides innovative components of expert systems based on business rules or existing neural networks in the market.

The most important are:

  • SIO is able to emulate the analytical capacity of an expert in any field of activity.
  • It is designed specifically to work in real time and with large volumes of data.
  • Able to learn from real-time data updating defined behaviour patterns.
  • It can detect behaviour not known to the user.
  • To have the ability for early detection of abnormal behaviour through analysis of trends by using patterns.
  • Provides a natural language explanatory component for each alert. In this way, the system provides auditability decisions.
  • Each alert has an associated quantitative level of severity.
  • The system can use physical models wherever necessary to achieve a higher quality solution.
  • The system combines the observation of different sensors that feed it by providing a hierarchical structure of data interpretation

 


Success Cases

 

AIA have been successfully applied in different environments:

  • Detecting fraud in payment means
  • Money Laundering Detection
  • Interpretation of alarms in the grid
  • Interpreting alarms in Hospital ICU environments
  • Interpreting energy bidding strategies of the agents of generation in the liberalized electricity market
  • Detecting cardiac abnormalities