This site provides a list of sensors of various types that generate streams of data. In some cases, for example with accelerometers, the data stream will be fast, large and complex. Even with simple data such as temperature readings, making business decisions based on the data can be complex. In all cases, there’s a stream of data that requires Time Series Data (TSD) analysis.
There’s a new article at IoT World Today explaining how TSD analysis is a new frontier. Developers need to learn data science and in some cases AI deep learning to make sense of the data.
The cost of adding sensors is low, and the required infrastructure is now ubiquitous. The major challenge is the availability of expertise to realize value from the data streams
Care needs to be taken to collect data at the right rate, fast enough to capture the required features in the data but not too fast so as to overwhelm storage and processing. When there’s too much data, edge processing needs to be considered which also allows for faster alerts and notifications than if the data is processed at a server.
There’s also sometimes need to mix and match data from different sensors for analysis and display. There often needs to be a pragmatic balance between accuracy and speed of analysis.
With so much more data to digest, the bottleneck in the process really becomes the engineers and computer and data scientists that used to devise, test and implement algorithms largely by hand-using high-level languages