Often the most useful data is not the direct sensor data, but information that is calculated or inferred from combinations of that data by an algorithm. An example is a real-time locating system. The raw reader data may provide strange results: a specific item seen in the back room and at various points while being brought to the store floor suddenly disappears from view without a trace. In fact, we can infer that the item is simply blocked from being read. Algorithms to decipher the meaning of low-level data are developed by experience in environments over time to create better and better higher-level inferences and intelligence. Data may also be federated from multiple sources to provide a broader view. For example, in an office building, data from the elevators, lighting systems, heating and A/C, security/surveillance, and other systems may be combined to provide a higher level of intelligence about the building. This may also include data from external sources such as weather, traffic, what time a big game is letting out, or other external events, to decide when to pre-heat, pre-cool, or take other actions in the building.
Enjoy curated articles directly to your inbox.