OSIsoft Shines a Light on Dark Data

May 1, 2018
Leveraging its core data management technologies as strategic repositories for decades of industry data, OSIsoft widens its aggregation and analysis scope with an edge-to-cloud approach.

As the supplier of one of the most widely used data historians across industry, OSIsoft finds itself in a sweet spot amid the upswing in attention around industrial data acquisition and analysis. Somewhat surprisingly, however, the company has been comparatively quiet as interest in the Industrial Internet of Things (IIoT)—and all its accompanying focus on data—surged over the past few years.

That appears about to change.

At OSIsoft’s latest annual user event—now dubbed PIWorld to reflect the event’s expanded attendance beyond operations engineers to include more IT personnel, data scientists and executives—the company focused heavily on its developing role in what it calls pervasive data collection.

When it comes to data, said OSIsoft CEO Dr. J. Patrick Kennedy, “The more people that see it, the more valuable it becomes.” This viewpoint drives OSIsoft’s work around pervasive data collection and underlies its efforts to increase the visualization and interaction capabilities of its PI System—a group of software and server products for management of real-time data and events.

OSIsoft characterizes its pervasive data collection work as shining a light on “dark data,” or data that is not currently being collected and analyzed but could be used to provide more holistic operational insights. The term dark data is also used by OSIsoft to refer to data that is being collected, but gathered in silos so that it is not aggregated with other data.

Chris Nelson, OSIsoft vice president of engineering, said, “We’re seeing a plethora of new sensors and devices being deployed for information-gathering purposes. And these devices are increasingly transmitting complex data types—not just time stamps.”

However, many of these new devices are often set up in ways that bypass traditional control networks for ease of data transmission. Likewise, data collection via Internet of Things (IoT) platforms can often bypass IT networks. Nelson said these bypasses mean that a lot of data go directly to the cloud for direct analysis—avoiding aggregation and thereby creating data silos.

To address this, OSIsoft is “extending the PI System to the edge and into cloud through our pervasive data collection vision,” Nelson said. This data collection vision is aimed at illuminating all the dark data.

One component of this extension will be OSIsoft’s Cloud Services Platform for data science, which will be available as a software as a service (SaaS), Nelson said. This platform will “maintain an operations data ecosystem that connects users to best-in-class analytics,” he explained. “The focus of this [platform] is for remote operations monitoring and to reach the dark data that used to be too costly to reach because of a lack of IT resources.” The platform can also be used to aggregate information from data silos.

Nelson added that the Connected Community aspect of the Cloud platform will allow users to unlock their data—at their discretion—for collaboration with partners and vendors. The platform will also feature Data Science Enablement for users who want to tap into PI System operations data for data science and machine learning initiatives.

As for OSIsoft’s extensions into the edge of the industrial network, Gregg LeBlanc, vice president of product management at OSIsoft, said the company will be supporting IoT sensing projects with its PI System Edge Data Store. This will support handling of data from sensor devices deployed on technologies ranging from Raspberry Pi to gateways.

The idea with this approach is to allow users to more easily capture data for input into their PI System, said LeBlanc. Some of OSIsoft’s customers, like those in utilities, “have widely dispersed assets, so they are looking at deploying these local functionality sensors with no moving parts and low resource requirements” to gather more data for analysis, he added. Plus, automation and device vendors are also looking for ways to enable products to deliver their data into a PI System or cloud service.

OSIsoft’s Cloud Service Platform and Edge Data Store are currently being tested at customer sites. The company expects to release them in the fall of 2018.

Two other key aspects of OSIsoft’s 2018 plans include the release of PI System Health and PI System Directory. “With these additions, users can spend more time using the data rather than monitoring the system,” said Nelson. PI System Health will monitor the health of data flow on the network while PI System Directory automatically detects PI System inventory.

Ultimately, it’s OSIsoft’s PI Vision, which was launched last year, that will deliver the information gleaned from all of the collected data to the users who need it. Though PI Vision is a relatively new product (it’s a 2017 update of OSIsoft’s PI Coresight product), Nelson said that OSIsoft’s new data collection vision necessitates “a major upgrade that will allow for greater mobile access to PI Vision and to deliver an app for it through which new features can be delivered.”

This upgrade, according to Nelson, will be an easy one for current users to implement and will also include a PI ProcessBook migration tool. ProcessBook remains a widely used visualization tool for PI System that predates PI Vision. This migration tool will reportedly allow users to migrate ProcessBook to PI Vision at their own pace.

Companies in this Article