Roger Larson October, 23

Choosing the Right Industrial Data Analytics Tool

Most manufacturing organizations have data in process historians or more traditional databases that can provide significant insights if analyzed with the right tools. Software tools that specialize in reporting, dashboarding, analytics, and/or visualization can provide valuable information to improve operational efficiency in manufacturing and processing environments.

There are a myriad of tools available today for industrial analytics. Some of these tools come from traditional automation software providers, but many come from companies that we are more used to seeing in the office. Here are a few products we have run across that provide an idea of what is out there, and hopefully will provide a starting point for further exploration.

Licensing Models

Some of these tools exist on-premise, some are be cloud-based and accessed via web browser, and some are a hybrid mix of the two. The main differences are location, access, and licensing model. In particular, on-premise software is usually a capital expenditure (license-based model) while cloud-based software is usually an operating expenditure known as “software as a service” (SaaS, subscription-model).


Industrial Analytics

Seeq is a tool for advanced analytics of process manufacturing data that is excellent at pattern recognition, predictive analytics, reporting, and dashboarding. It is capable of ingesting data from SCADA systems such as Ignition and Wonderware, and can be used as an on-premise or cloud-based analytics system. It is possible to analyze data from enterprise-level systems in Seeq and obtain high-level analyses and forecasts (e.g. mechanical failures, cost inefficiencies, process optimization).


ThingWorx Analytics is a product by PTC that can analyze industrial IoT data. It is able to collect and retrieve data from devices or other sources for real-time analysis. It has machine learning and automation capabilities, can do predictive analytics, anomaly detection, and remote monitoring.


Generalized Analytics

Statistical Analysis System (SAS) has a rich history in analytics as one of the oldest software packages available. It can do almost any statistical analysis, has all machine learning functionality, can be used in nearly all industries, and can connect to IoT devices or most data sources available. It even has its own language (SAS language) for other kinds of custom connections, transformations, and analyses.


Cloud-Based Analytics Suites

Microsoft Azure has various software packages that can analyze data. Any collected data can be ingested in real-time via Azure Event Hub, and transformed or analyzed via Azure Stream Analytics. Azure Time Series Insights can be used for root-cause analysis and anomaly detection, and Azure Machine Learning studio can be used to do any kind of advanced analysis on any data source. In particular, ML studio can take data directly from Ignition SCADA, run analysis externally, and return data back into Ignition via web calls for visualization or use by operators.


Amazon Web Services (AWS) has various analytics tools equivalent to Azure. Streaming data can be analyzed in real-time with Amazon Kinesis, or operational analytics with Amazon Elasticsearch Service. Machine learning can be used on any data source via Amazon SageMaker for any advanced analytics such as forecasting and optimization of processes. Amazon QuickSight allows for dashboards and visualizations of analyzed data from connected data sources.


Dashboarding and Visualization

Microsoft Power BI is a dashboarding and analytics tool that can connect to various data sources within Azure, or any external data source, and can create useful dashboards, forecasts, and visualizations. Tableau and Zoho function similarly in their dashboarding and analysis capabilities.


Custom Analysis Software

Sometimes, existing software tools might not be the right fit for the analytics problem due to integration issues. Custom analytics software can be written in programming languages such as Python, R, Java, and Scala, that can then be deployed as a service for any solution. The downside of this is long development time and the need to outsource or have a software or data engineer on hand to design and deploy the custom program solution to the analytics problem. The major benefits of custom programs are that development is a one-time cost, licenses are irrelevant as the company will own the software, and the solution can be tweaked or re-used as needed.

Dig a little deeper into Industrial Data Analytics in this blog post.



Sign up to get the latest from Vertech delivered right to your inbox.