During "commissioning" – the process by which all of the nodes in a facility are brought online and configured into the server – data from the hundreds or thousands of sensor nodes flows into an advanced AI engine which forms the core of the management network. As you can imagine, the complexity of these environments requires sophisticated processing power to address the real-time requirements of thousands of variables changing many times a second.
Our approach (the details of which are proprietary and patented), is to develop a comprehensive data model of all system inputs and outputs – we treat the environment as its own ecosystem, as it were – feeding the AI engine to handle the decision making at a level out of the reach of most people.

The system also configures to the cooling and air-handling units in the facility, using BACnet or other protocols to communicate with the units. This provides an essential component of the closed-loop approach utilized by Vigilent – measuring and monitoring is a good and necessary start, but it's essential and imperative to apply intelligent and automatic control to manage the complex interactions among heating and cooling sources, and airflow within the facility.
The system is adaptive and dynamic, reconfiguring itself and its responses as its understanding of the feedback systems in the environment develop. As commissioning gets underway, the system begins to perturb the environment, provoking responses and analyzing the results to determine the interplay between each and every sensor node and cooling unit in the network.
The profile that is created forms the basis for a model that is updated regularly and constantly as site conditions change and evolve, allowing sophisticated what-if decisions and simulations to be explored to determine the optimal configurations and control for a facility.
Artificial intelligence – Advanced technology addressing issues as reasoning, planning, learning, perception and knowledge representation.
Neural net – An AI technology used to solve complex intelligence challenges using autonomous and adaptive learning approaches.
Transduction – A form of machine learning in which inferential reasoning is used to move from observed training cases to specific cases; it is an advance over earlier inductive techniques.
Once only the province of fiction, serious work on modern artificial intelligence began in 1956, building on the pioneering work of Alan Turing in programmable logic a decade earlier. In the last ten years remarkable progress has been made in applying AI technology to practical issues. the Vigilent energy management systems make extensive use of AI to deliver the sophisticated, real-time control that is almost impossible through any other means.