Category : Uncategorized
The key topics discussed in this technical brief :
Why programmability is desirable at edge or OT site
Programming analytics at edge through IT center
Distributed plant assets but centralized control
Going beyond static set-point events to AI at edge
Sayantek’s eRED is a software product for edge analytics and that extends open-source Node-RED platform for Industrial applications. Today IIoT practitioners describe a big divide between OT and IT. Edge is OT or assets in a plant, while IT is infrastructure in the cloud or in an enterprise datacenter or at a headquarter central monitoring system. IIoT is essentially connectivity of plant assets (OT) to IT systems. At Sayantek, we want to go beyond OT to IT connectivity and enable applications or analytics at OT infrastructure that is unified with IT applications. The purpose of analytics is to extract meaning and core information from raw data at edge or plant where data originates. This may be events or something more like statistical heavy lifting at the edge.
eRED requires two components- the first component on a plant site called “eRED-Edge” runs on an edge computing device like an industrial groov device from our OEM partner Opto 22 or an on-prem server. The second component within cloud or enterprise data center called “eRED-Controller”on a server, integrates flows across multiple eRED-Edge devices with IT business or predictive analytics applications. eRED-Edge has library or nodes in Node-RED that we have built for various analytic components at edge – data acquisition, data ingestion, cleansing of data, data rationalization and processing. The application infrastructure at edge is ingesting into time-series database and python backend for analytics. We have simplified building an analytics flow in Node-RED with chaining of nodes with minimal programming background.
eRED-Controller resides in central IT ( cloud or enterprise datacenter) and combines eRED (Node-RED) flows across edge compute devices as a single Node-RED flow. An important capability of eRED is life cycle management of applications, that is, build, deploy and drive new services or analytics to edge devices or on the OT side that are part of a unified flow. IT and OT are usually separated by a firewall, and we have built secure connectivity across firewall. Another use case is remote edge or plant infrastructure that has limited network or cellular connectivity. The use of edge analytics combined with distributed flow management at orchestrator results in substantial reduction in usage of WAN links. eRED solution thus makes the entire path from plant to cloud or datacenter programmable.
We have partnered with Opto 22 to enable the next generation of plant connectivity and edge analytics. The integrated eRED with SNAP PAC enables data acquisition from tagged plant assets as a node in Node-RED. This enables a direct pathway of plant assets to IT in cloud or datacenter, unlike traditional way of using gateways or middleware. eRED analytic stack has been integrated with software in groov device for edge computing and mobile visualization. The combined solution is capable of driving a new class of Artificial intelligence at edge based condition monitoring and predictive analytics. This is clearly well beyond set-point based events that are now deployed in industrial process, control and monitoring systems. Our combined solution can be applied for new forms of monitoring, control, optimization, operational efficiency and realize the potential of IIoT to power new business models.
We have an AI based anomaly detection application to monitor quality of water and waste-water. The core of our solution is based on award wining US EPA approved analytics engine to monitor chlorine, pH, nitrite, ammonia and others. The same application can be applied to a range of other problems in IIoT.