Collect, transport, and distribute data in real-time. As data moves from one stage to another latencies start adding up.
Sensors have always been able to produce data at high rates. Acquisition of high frequency data was always the bottleneck. Combine that with transporting this high density data over low bandwidth networks that are intermittent at best, further complicates use of this valuable resource. With the available computational capabilities it is now possible to listen to everything the sensors have always been telling us.
Analytics needs quality data. Streaming analytics require systems that can ensure streaming data quality. High frequency data comes with high frequency noise. In complex operations such as manufacturing assembly lines and drilling for oil & gas, one of the keys to identifying noisy or spurious data is to be context aware.
Use high density data to not only extract insights but also turn insights into decisions that can be executed in a closed loop system. Component to system wide automation is a journey that can only be undertaken with the right combination of data, software, and hardware.
Drawing insights from streaming data in the context of metadata is the only way to automate decision making.
Deploy on a range of architectures from an ARM processor to an HA cluster, a 1U box at the rig to a virtualized server farm in the cloud. Given the volumes of data being produced, edge computing is an essential and necessary approach to utilizing this resource.
Monitor data producers, underlying infrastructure and software components to ensure continuous data collection and distribution. With so many moving parts in this complex supply chain of data, tools to monitor the health of the producing sensors, the SCADA systems, and the networks, are an absolute must have. Without the blessing of a well appointed monitoring system all data is suspect and cannot be used in any meaningful decision making.
Continuously improving both technical capabilities but also communication techniques to deliver quality results faster is what makes a team and an organization agile. It's not about being the first to market but being the quickest to react to the operational exigencies of today's world of IoT, that sets an organization apart.