Getting the right tool for the job

20 May 2018

Michael Risse gives advice on evaluating advanced analytical solutions to help ensure you have a tool that meets the very particular requirements of the process industry.

New Industrial Internet of Things (IIoT) deployments across the process manufacturing sector, along with a corresponding need to derive insights from the resulting data is driving the need for data analysis solutions.

The results will only occur when the selected analytical tool closely matches the application requirements. The best way to achieve this match is by carefully formulating a set of questions for your chosen software vendors and examining their answers.

First, however, it is important to correctly understand the definition of advanced analytics. Put simply, advanced analytics software enables process engineers and experts to do the following, without the need for assistance from data scientists and IT personnel:

• Create a cleansed, focused data set for analysis through assembling, aggregating or ‘wrangling’ data from various sources including data historians, offline data, manufacturing systems and relational databases.
• Investigate process data using self-service tools to rapidly analyse alarm, process or asset data for ad-hoc or regular reporting.
• Capture efforts for collaboration with colleagues through documentation of work
• Avoid locked/hidden IP within spreadsheets and formulas.
• Publish or share insights and reports across the organisation to enable data-driven action, or to enable predictive analytics on incoming data.

Many advanced analytics solutions claim to offer some or all these things, with the ultimate goal of closing the gap between data and insight. Here are four key questions, which once answered, should help clarify which solution will best meet your requirements.

A process industry focus
If your application is in a process industry such as food & beverage, oil & gas, pharmaceutical, power, water/wastewater, mining, or chemical, you need an advanced analytics solution specifically designed for these types of applications. Trying to fit a general-purpose solution into a process industry application is possible, but will require a great deal of time and effort on your part.

With this in mind, it is important to ask whether the tool is designed specifically for process industry applications, and whether it can handle time-series data and be easily used to solve intricate process manufacturing problems.

Historians are typically used in the process industry to collect data from many different plant and facility areas, and to store it in time series databases.

Sources of process industry data can include PLCs, HMIs, MES software, and asset management systems. All of these devices produce data at prodigious speeds and volumes, and at uneven intervals which can easily confound conventional relational databases. Most process data, typically, needs to be cleansed to be useful as much of the data lacks the context to make it useful – a problem that is further compounded when analysing data from multiple sources.

An advanced analytics solution suitable for process industry will need to work time series data and should include connectors to all leading process industry historians to allow the solution to handle, display and navigate time-series data.

Process experts
The key to positive outcomes is empowering internal process experts so an important question should be, is the solution designed to be used by process engineers and experts, or by data scientists and IT personnel?

If it’s the latter, then your process experts will not be able to interact directly with the data of interest to gain insights, but instead will always go through others who must possess a high level of specialised IT expertise. This creates a situation where interactions and iterations take too much time to be useful.

For example, a process expert identifies a problem area, and asks their IT expert to examine the data for possible clues. The expert provides what he or she thinks is the right answer a few weeks later, but this is probably not what the process expert was looking for.

With the right advanced analytics solution, there will be a direct interaction between process experts and data, giving quick answers to questions using the iterative method required to solve most difficult process problems.

Focused on problems
?The focus should be on solving common process industry problems, not on the underlying technologies employed in the advanced analytics software. In too many cases, the technologies – big data, predictive analytics, machine learning, cloud computing, etc – have eclipsed the desired process improvements. Rather than discussing why one should adopt a particular product, the conversation focuses on what technology to use, often with more enthusiasm for the technology than the benefits.

The goal of any data analytic solution should be to improve yields, margins, quality and/or safety. Any modern solution should be drawing from recent technology innovations to accomplish these outcomes, without requiring expert assistance or knowing exactly how these underlying technologies work.

The world of big data, predictive analytics, machine learning and cloud computing really needs to be turned inside out – from a technology-centric revolutionary approach to a user-focused and problem-solving evolutionary approach.

The right advanced analytics solution won’t ask you to do a lot of work to adopt specific innovations. It will harness innovation on your behalf by using technology advancements to deliver concrete benefits specific to process manufacturing via a modern, cloud-enabled and browser-friendly application experience.

Data interaction, not data movement
The advanced analytics solution should not require you to move, duplicate or transform data. A better approach is to create an index on top of your data sources so your experts can interact with the data in a structured way, while leaving the data itself in place. Your experts can then search the data like they would with Google, and quickly and dynamically add context across data sets.

When the advanced analytics solution connects directly to historians and other data sources, process exerts can contextualise without getting IT involved, without creating data lakes, and without duplicating or transforming the data.

The advanced analytics solution needs to connect to your live data – as it is and where it is, of any size and any type – so your expert can interact with it directly. Your experts will then be free to traverse the system, ask questions on the fly and layer multiple data sources on top of each other in a single view, even when many data sources are involved.

All work is captured for future use and reference, so engineers don’t need to start all over again if their original question doesn’t prove out. Results can be easily shared across the enterprise, enabling collaborative efforts.

With the right advanced analytics solution, process experts can quickly bridge the gap between data and insight. This will make difficult problems easy to solve. The result is faster insights leading to better yields, margins, quality and safety outcomes.

Michael Risse is vice president & CMO at Seeq Corporation.

Contact Details and Archive...

Print this page | E-mail this page