Derive Insights from Machine Data with InfluxDB
Jessica Wachtel /
May 31, 2023
The panel discussion “From Machine Data to Business Insights, Building the Foundations of Industrial Analytics” discussed modern methods and benefits of deriving insights from machine data. InfluxDB Developer Advocate Jay Clifford explained the trend now is to “allow the builders to bring the Lego blocks and build them together how they see fit. ” Companies don’t need to tolerate vendor lock-in requirements and obscene prices because modern solutions consist of open-systems with flexible technology.
When building a full data capture-to-analysis solution for a modern factory, it’s important to keep a few things in mind. Only use components built for scale from the ground up, no matter what size the operation is today. Scale in this case means there’s space for exponential growth, from a single machine to over 10,000 machines overnight, without changing anything. It’s also important to have clear goals in mind when building a datapoint monitoring solution. Return on investment through increased operational efficiency or overall equipment effectiveness are the most common desired outcomes. For this, machine data-driven insight is necessary. This takes place in three stages – data collection, storage, and analytics.
The remainder of this post introduces an open solution for collecting, storing, and analyzing machine data. It includes Kepware for data collection and InfluxDB for storage and analytics.
Deployment Facilitation – Balena
Balena is secure, container-based technology that enables engineers to develop, deploy, manage, and scale IoT devices. Balena facilitates the deployment of Kepware and InfluxDB at the edge and monitors the edge gateways running the software. Balena uses Docker containers which make for scaling at ease since developers can update any dockerized image at scale.
Data collection – Kepware
Kepware, represented in the webinar by Kyle Carreau, is a scalable PLC, IIoT connectivity solution. Kepware connects to devices with different protocols, and provides users with a single source for industrial data. Working with a flexible solution like this is important because data collection is one of the biggest hurdles in this process. Though OPC-UA, MQTT, and Modbus are now standardizing the industry, capital equipment has a lifespan of 20-25 years. This means the lack of standardization between industrial protocols still exists inside modern factories. Carreau said, “customers who don’t solve that connectivity challenge early in the process, that project fails every time,” in regard to implementing a scalable, open data collection system.
Storage – InfluxDB
Kepware sends the data from machines to InfluxDB. InfluxDB was purpose-built to handle time series machine data at scale. InfluxDB is currently crossing a milestone when it comes to data ingestion due to customers’ new preference for higher data resolutions. Now data is sent to InfluxDB in milliseconds, sometimes nanoseconds, depending on the types of machines sending data. The level of sophistication continues to raise the bar and InfluxDB is there to meet it.
Edge and cloud computing
Both edge and cloud computing have their place inside the modern factory workflow — one isn’t a replacement for the other. Oil refineries use an InfluxDB OSS edge node inside the well pump. This allows site engineers and operators to make split-second decisions when needed. Site engineers then send aggregated data to InfluxDB Cloud hourly. Data scientists and engineers can then access the data to perform further analysis and modeling on data from oil sites across the globe.
Because edge computing and cloud computing aren’t drop-in replacements for one another, InfluxDB has an edge node and cloud offerings. The edge provides customers with the opportunity to work next to the data and offers a faster reaction time. The cloud provides more opportunity for work with larger data sets, ease of use with the introduction with SQL, and a database engine built on Apache Arrow.
Analytics - InfluxDB
Once the data is in InfluxDB Cloud Serverless, it doesn’t have to go anywhere else. InfluxDB is feature rich and integrates with machine learning, artificial intelligence, and statistical analysis tools necessary to discover connections between systems that aren’t apparent without technological intervention. Customers can use InfluxDB’s flexible technology to build the data-driven insights that work for their business.
One customer uses InfluxDB analytics to monitor the vibration metrics on multiple CNC machines. When the data from one machine begins to deviate from the others, operators use dashboards powered by InfluxDB to get a direct view into the deviation. From there, on-site teams determine what further actions to take.
The creation of digital twins is an example of a more sizeable use case. Large manufacturing firms create digital twins from highly granular machine data and InfluxDB analytics. PLC engineers use the digital twins as the training grounds to test standard operations, problem scenarios, and work out optimizations. This results in more optimized code without having to test it on physical PLC machines that are in production.
The technology needed to execute customers’ goals exists, but it requires time series machine data. Building a solution that works not only for today but also for tomorrow is of utmost importance as well and that means choosing the right technology partners. Kepware is a flexible data collection software that connects to PLCs and IIoT devices that send data to InfluxDB. InfluxDB is a scalable, purpose-built time series database that handles all your storage and analytics needs.