Industrial Internet Platform

The Power Of Predix: An Inside Look At How Pitney Bowes Is Using The Industrial Internet Platform

GE opened Predix, its cloud-based operating system for the Industrial Internet, to all users at Mobile World Congress in Barcelona on Monday. As of now, companies of all stripes can start using it to write industrial applications and make their machines and factories run better.

But a handful of businesses have had access to a beta version of the system since last fall. Roger Pilc, executive vice president and chief innovation officer at Pitney Bowes, has been developing on Predix for about six months. We talked to him about the experience. Here’s an edited version of the conversation.

GE Reports: How are you using Predix?

Roger Pilc: We leverage physical and digital technologies to help our clients drive business outcomes. We have an enterprise mail business that develops, sells and services very large machines called inserters. The machines help companies like banks and healthcare providers put exactly the right piece of mail inside exactly the right envelope. The scale and speed of the work is tremendous. For one bank alone, our machines helped assemble 900 million mail pieces in one year, just to give you a sense of scale. Most of these machines produce over 20,000 letters per hour and the requirements for accuracy and precision are astronomical. You can’t get it wrong.

This is not an easy feat. We need visibility into the machines to avoid outages and fix the machines before they break down. We also need to understand the root cause of any problems very quickly. Predix is helping us and our customers do that. This is something we didn’t have before.

GER: Is Predix meeting your expectations?

RP: We’ve been pleased. The data is successfully coming off our machines and through our security systems up into the cloud. The applications running on top of the cloud have been showing our field service teams and our customers the data visualizations we have designed together. So in terms of technology and functionality, things have been going very well.

GER: Besides outage prevention, what else are you using Predix for?

RP: So far I’ve mentioned one application, but our customers have a hierarchy of needs. Reducing downtime is the foundation level. The next step is improving productivity.

Predix can do that by using some of the same data from the same sources as in the first example, and applying an additional layer of analytics and different applications. They allow our customers to manage uptime, and also productivity in terms of output and how it relates to the mix of the types of machines, applications and operators in the factory. It gives our customers the analytics to optimize those machines even more.

GER: You talked about reducing downtime and improving output. Anything else?

RP: Helping our clients grow productivity revenue is the third and most advanced level of our use of Predix. When we talked to our customers on the most senior level, they were ultimately most interested in driving more revenue and more profit out of their locations. For that, you need to go even higher with capacity planning and job scheduling. Predix allows you to plan production in a more informed and data-driven manner against a whole fleet of machines. It allows you to place the right applications and analytics on the right machines at the right time, and essentially foresee what will happen in the future.

Here’s another way of looking at it. There’s descriptive analytics, which allows you to describe and visualize what’s happening; there’s prescriptive analytics, which is based on what’s happening — the root causes behind it — and gives you the prescriptions for remediating it; and then there is predictive analytics, which allows you to use data to be able to foresee what will happen and to be able to act in advance of it to optimize outcomes even further.

“The scale and speed of the work is tremendous,” says Roger Pilc of Pitney Bowes. “For one bank alone, our machines helped assemble 900 million mail pieces in one year, just to give you a sense of scale.” Image credit: Pitney Bowes

“The scale and speed of the work is tremendous,” says Roger Pilc of Pitney Bowes. “For one bank alone, our machines helped assemble 900 million mail pieces in one year, just to give you a sense of scale.” Image credit: Pitney Bowes

GER: Have you worked with other Industrial Internet software platforms?

RP: We certainly have experience with other leaders in the industry. We are a large customer of Amazon Web Services (AWS) and we also work with another industrial-strength hosting company. We evaluated Predix thoroughly and were very satisfied with the technology architecture, the level of scalability and the ability to ingest the very large quantities of data that often come from our types of machines. We are also very comfortable with the level of security, which is very important to us and to our customers.

Next, we looked at the actual applications. Here, Predix for us is more like than AWS. The critical element was the layer of asset performance management applications running on top of the technology and the analytics that informed those applications. The application layer where the data is used to drive the performance of our industrial assets was something that we viewed as the industry leader. Plain-vanilla application-hosting companies do not have that layer.

Finally, and this is also very significant, these Predix-enabled applications are helping us provide more value to our customers, get to better business outcomes and also transform our own operations. We are using Predix to transform our own service organization from a break-fix model to an increasingly consultative one informed by data and analytics. We want to engage with customers in a way that’s more directly related to business outcomes.

GER:That sounds similar to what GE is doing inside its own factories.

RP: We placed value with GE because they not only built the data and analytics platform, but also because of the journey they’ve been on for the last several years. They’ve been using data from their own machines, doing the data analytics and then ultimately evolving their own services organization in the exact same way as us. That was an important element to us. We speak regularly not just about the technology and the applications, but also about this digital-industrial transformation.

GER: How has it been working with GE so far?

RP: It’s been very good. First of all, I think that our two companies have an amazing cultural fit. We both care a lot about our customers, we try to do the right thing the right way and we both have a high bar in terms of the quality of our engineering. We also share GE’s strong focus on innovation, meaning that we try to meet customer needs very quickly by doing things differently and introducing new physical and digital technologies.

GER: What are your plans in the future?

RP: We’re a $3.6 billion company with 1.5 million clients. In addition to the mailing machines, which are a relatively small part of our business, we are leaders in software and analytics around geospatial and location intelligence and customer information management. We are now using Predix to develop apps around those services and offer them to customers through the Predix app store.

Any Predix customer can use these capabilities, from fleet management businesses to drilling companies. Location management is a fundamental enabler. I’m sure other companies will be writing apps leveraging these capabilities as well. This is the power of the Predix ecosystem.

Want more stories like this delivered to your inbox?