The solution kicked off with the installation of on-site OSIsoft PI infrastructure, in order to collect:

• Operational data from SCADA systems, 

• Maintenance and audited record data from SQL databases, 

• Laboratory data from wastewater treatment and sewer testing locations, and 

• RESTful API data from smart water meters.  

The data from these systems, and others, were combined into a variety of meaningful “asset hierarchies” – data models that broke the individual data streams down into their respective place in the real-world infrastructure of the businesses. Some of these asset hierarchies represented all of the customer water meters in a given area; some represented pumping stations and their pumps; some represented chemicals and contaminants at different testing locations; but all of them were used in the pursuit of our deliverables. 

Once the data was collected and modeled correctly, we continued to work with subject matter experts on ways to transform it. For some areas, this meant automating the production of reports on a scheduled interval. For others, it meant adding calculations on top of the raw data – to detect trends in performance that were previously hidden to operators and to management, or to replace analyses that took many days to compile manually in spreadsheets. 

The critical piece in this process was the just-released (at the time) “PI Cloud Connect” software, which was leveraged to take data from each organisation up to a cloud-based PI server, hosted in Microsoft Azure. It was here that many of the transformations took place; having a centralised location allowed us to define the calculations just once, but roll them out to both customers. 

For some deliverables, automated spreadsheets were the endpoint. But for most, we wanted to hand over appealing, information-at-a-glance visualisations to our customers. Combining PI Vision with our asset hierarchies allowed us to do just that, with bespoke dashboards created for pump performance, rainfall and infiltration, regulatory reporting, smart water meter and wastewater treatment applications.

To complement the time-series nature of the PI Vision dashboards, a cloud-based ESRI ArcGIS and GeoEvent server were added to our technology stack, and relevant assets were given latitudinal and longitudinal co-ordinates in the relevant asset hierarchies. When combined with the OSIsoft Integrator for ESRI ArcGIS, these assets, their operating statuses and real-time values became accessible on a map. This allowed for problematic stretches of sewer mains to be identified, and for customers with water usage far more than usual to be identified (and proactively notified, rather than getting a nasty case of bill shock two months later). And with both PI Vision and ESRI, the dashboards could be made available to any authenticated user, in any browser and on any device, anywhere in the world.