Please download the attachment "OSIsoft Q&A Feb2018 Webinar.pdf" below for the complete document.
Q: What are the most common challenges in implementing an enterprise-level PI System?
A: Bob: The two that I find the most challenging are data silos, people have these all their operations data in their own private data system and they don’t want to share it with anybody and, resistance to trying to integrate it all into one big system. The biggest challenge is how do I get all the source data collected. The other big one is the governance type of thing. Having that board, across the top, do these various things and these different groups work together.
Q: What’s the difference between the PI System and a data lake?
A: Heather: The key difference between the PI System and a data lake is the ability to store operational data at high frequencies for the life of the asset. The PI System is specifically designed to be able to store data at the appropriate interval – from milliseconds to daily intervals – whatever’s required to make the data meaningful. For example, for PID loop tuning or pressure transient analysis – you need high-frequency data. Also having that data online for the life of the asset makes it possible to access the data whenever it's needed. “Data Lake” options are not designed to manage this granularity of data for the life of the asset.
Q: Question for Bob: did you have market standard to help inform your tag standards? Any specific challenges with sites in different geographies that may have different rules to follow?
A: We went to English tag name standards instead of loop numbers. Tag names drill down from the top to the equipment and measurement. Site code – area code – process/equipment – measurement. We had to pick something that works well with the way AF handles templates.
Q: I was wondering if you had any common problems/roadblocks on the IT side when getting PI System data from the sites out to the enterprise networks? Any golden nuggets of advice in this area would be great
A: Bob: Working closely with the IT security and network/firewall groups is key. You need to have a good working relationship with them. There are lots of different ways to get the data through securely. Working with them will give you a workable solution for your environment. Heather: I totally concur with Bob. Working closely with IT from the beginning ensures a sustainable solution and both sides learn about any limitations and work together to find resolutions. It really helps to build that collaborative team approach.
Q: What is the architecture of your Enterprise System? Does the Enterprise PI Server talk to the site PI Servers or directly to the site data sources? Is the standard PI Tag name implemented at the Enterprise level only or at sites also?
A: A single enterprise level PI Server. Some people initially kept their site PI Servers, but shut them off when they didn’t use them anymore. There may be cases where someone needs to keep theirs and that is fine too. If the site server is a PI Server, then I use the PI to PI Interface. If site server is an “other” server then I use OPC-DA Interface and run it in parallel to their server. The tag naming standard is implemented at the enterprise level server only.
Q: Does the PI System have a cloud service where companies can store the data?
A: OSIsoft: OSIsoft currently offers PI Cloud Connect where the customer can share data between multiple PI Systems. Our question here is what are you trying to achieve. Visit our OSIsoft Cloud Services page here. This page gives you an overview of PI Cloud Connect, as well as a link to an FAQ. However, I highly encourage you to get in touch with your Account Manager.
- OSIsoftLearning - YouTube: PI Asset Framework (PI AF) related topics
- Have conversation with Heather and Bob