How to Get PI Data into Power BI

Getting your operational data from the OSIsoft PI System into a Power BI dashboard can feel like a major hurdle, but it unlocks incredible potential for analysis and decision-making. Seeing your real-time and historical process data alongside business metrics gives you a complete picture of performance. This guide will walk you through the primary methods to connect PI to Power BI, breaking down the steps so you can start visualizing your time-series data without the headache.

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

Before You Begin: Key Considerations

Before you jump into a specific method, it's wise to think through a few key points. Answering these questions now will help you choose the right path and save you from getting stuck later on.

• What version of PI Server are you using? Some connection methods, particularly newer ones, rely on more recent versions of the PI Data Archive and the PI Asset Framework (AF). Knowing your environment is step one.
• Do you have the PI Integrator for Business Analytics? This is OSIsoft's (now AVEVA's) dedicated tool for this exact purpose. It's the most powerful and scalable option, but it is a separately licensed product. If you have it, it’s almost always the best choice.
• What are your data requirements? Are you looking for aggregated data (like hourly averages) or raw, high-fidelity data? Do you need a live, streaming view, or is a daily refresh enough? Your answers heavily influence the best connection method. Large volumes of raw data can quickly overwhelm Power BI's import mode.
• What are your permissions? You'll need appropriate read access to the PI Data Archive and/or AF databases you want to query. Check with your PI administrator to ensure your service account or user account has the necessary permissions.

Method 1: The Official Route - PI Integrator for Business Analytics

The PI Integrator for Business Analytics is the gold standard for connecting PI data to platforms like Power BI. It's a purpose-built tool designed to efficiently extract, clean, and shape your complex time-series data into a simple, BI-friendly format (like rows and columns).

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

Why Use the PI Integrator?

Instead of hitting the PI Server directly with queries from Power BI, the Integrator acts as a middle-layer. You define a "view" of the data you want, the Integrator processes it, and then exposes it for Power BI to consume. This protects your PI Server from heavy query loads and does the tough data transformation work for you.

• Performance: It’s optimized for handling large volumes of historical time-series data without slowing down your operational systems.
• Data Shaping: You can pivot, average, interpolate, and aggregate data server-side, meaning Power BI gets a clean, ready-to-use dataset. This is huge, as reshaping time-series data can be very difficult in Power BI M or DAX.
• Ease of Use: The web-based interface is straightforward, allowing you to select assets and attributes from your AF structure visually.

Step-by-Step Guide Using the PI Integrator

The process involves configuring a data view in the Integrator and then connecting to it from Power BI.

1. Choose Your Data in PI Asset Framework (AF): Start by identifying the data you want to analyze. This is usually done by working with an AF structure that models your physical assets (e.g., pumps, turbines, production lines) and their associated data points (e.g., temperature, pressure, status).
2. Create a New View in the Integrator: Log into the PI Integrator's web interface. You'll create a new "View" and select the AF database and elements you identified. You can drag and drop assets or use search functions to build your initial dataset.
3. Shape Your Data: This is the most powerful step. In the view configuration, you can:
4. Publish the View: Once you've defined your view, you need to "Publish" it. The Integrator will give you a few publishing targets. For Power BI, the most common is an ODBC connection. As it publishes, the Integrator processes the data and stores it in its own internal SQL database.
5. Connect from Power BI Desktop:

The benefit of this approach is that you can set the Integrator to refresh the view on a schedule, and then simply schedule a refresh in the Power BI Service to keep your dashboards up to date.

Method 2: Direct Database Connection with PI OLEDB Enterprise

If you don’t have the PI Integrator, you can still connect to your PI data using PI OLEDB Enterprise. This tool allows you to query your PI Asset Framework using standard SQL commands. It presents the hierarchical AF structure as relational database tables that tools like Power BI can understand.

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

Who Is This For?

This method is more technical and best suited for users who are comfortable with SQL and database concepts. The performance won't match the Integrator for very large datasets, but it's a solid option for targeted analysis and smaller-scale dashboards.

Connecting with PI OLEDB Enterprise

1. Install the Driver: First, ensure you have the PI OLEDB Enterprise client driver installed on the same machine as Power BI Desktop. You can get this from the AVEVA support website.
2. Get Data in Power BI: In Power BI, click Get Data and choose OLE DB. Don’t select the SQL Server option, as this won't work correctly. This will open a window asking for a connection string.
3. Build the Connection String: This string tells Power BI how to connect to the AF Server via the OLEDB provider. It looks something like this:
4. Write a SQL Query: Unlike the Integrator method, you can't just pick from a list. You must provide a SQL query to get the data. You’ll be presented with a box to enter your SQL statement. This requires knowledge of how PI OLEDB Enterprise exposes AF tables.
5. Load and Visualize: Once your query runs successfully, the data will be loaded into Power BI, and you can begin creating visuals just like with any other data source.

Method 3: Flexible but Advanced - The PI Web API

For developers or those with a very specific, custom use case, the PI Web API is a powerful and flexible option. This is a RESTful API that allows you to query and retrieve PI System data over HTTP.

In Power BI, you can use the "Web" data source to call the PI Web API. This requires writing scripts in Power Query's M language to handle authentication, make the API calls, and parse the JSON response that is returned. While this offers the most control, it is also the most complex path and not recommended for beginners. You have to manage paging for large result sets and handle data transformation within your M scripts.

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

Best Practices for Visualizing PI Data in Power BI

No matter how you connect, visualizing time-series data has its own set of rules. Here are a few tips to get the most out of your setup:

• Use Aggregates When Possible: Avoid pulling in massive amounts of raw, second-by-second data. This will slow your Power BI report down to a crawl. Use the aggregation and filtering options in the PI Integrator or your OLEDB query to only pull the level of detail you need - like hourly averages or daily totals.
• Embrace the Line Chart: The line chart is the king of time-series visualization. Use it to plot your process values over time. Add markers for key events or annotations to highlight important periods.
• Leverage Hierarchies and Slicers: If you import data that includes your AF hierarchy (e.g., Region > Site > Asset), use slicers in Power BI to allow users to easily drill down and compare performance between different assets or locations.
• Start Simple: Don't try to build the ultimate everything-dashboard on day one. Start by answering one or two specific business questions. For example, "What was the average energy consumption of Line 3 last week?" Once you solve that, build from there.

Final Thoughts

Connecting your PI System to Power BI opens up a powerful new avenue for operational analysis, blending your process data with business intelligence tools your whole organization can use. We’ve covered the primary methods, from the streamlined PI Integrator to more direct OLEDB and Web API approaches, each with its own place depending on your available tools and technical comfort level.

While connecting specialized systems like PI often requires these technical steps, we know the struggle of data wrangling isn't limited to operational data. For pulling together marketing, sales, and e-commerce analytics - from sources like Google Analytics, Shopify, and Salesforce - we built Graphed to simplify the entire process. You can connect your sources in a few clicks, ask questions in plain English, and get live, interactive dashboards in seconds. It’s about spending your time on insights, not on figuring out connection strings and cleaning data.