Webinar : Contextualized Data : September 14, 2022 09:00AM Pacific Time
Register
Many plants built strong operational frameworks on PI Asset Framework under OSIsoft's legacy. AF brought structure to raw time series data and helped teams organize equipment, KPIs, and events, but at a cost that was pushing the boundry of affordability. Now, under the AVEVA brand, the cost and complexity has increased severity without seeing corresponding benefits. To further complicate the situation, AF is being forced even tighter inside a closed ecosystem. As systems diversify, that dependency slows growth and makes it harder to bring new data sources into a unified model.
A Timebase Historian plus InfoHub architecture keeps the strengths of AF but removes the licensing, storage, and integration constraints that come with AVEVA. You keep the intent of AF and gain a modern foundation built for scale.
This creates a complete, site-wide architecture for historization, data modeling, event frames, and KPI governance with no tag limits and no growth penalties.
This pricing shifts the economics of industrial data. Tag-based expansion stops being a barrier. Teams build full historical archives and consistent information models without incremental licensing.
In November 2023, the City of Lakeland, FL, approved a sole-source agreement with AVEVA to replace their legacy eDNA Historian with the modern AVEVA PI System. The 3-year subscription contract was valued at $325,710.00, resulting in an annual cost of $108,570.00. This agreement was structured under the AVEVA Flex model (more on this later), purchasing a specific pool of "Flex Credits" that allows the utility to run the PI Server, PI Vision, and associated interfaces.
The contract reveals that the pricing was locked in to replace the end-of-life eDNA system, and it explicitly references the credit-based "Top-Up" mechanics if the utility exceeds its burn rate. The purchase was justified as a sole-source procurement due to AVEVA's proprietary ownership of the software and the necessity of running both systems in parallel during the migration.
A utility of this size typically is historizing between 25,000 and 50,000 tags, no more. This could be accomplished with a single instance of Timebase Historian running on a 4 core, 16 GB server.
In recent years, AVEVA has fundamentally shifted how it sells its flagship industrial software, including the ubiquitous PI System. Gone are the days of straightforward perpetual licenses and annual maintenance fees. In their place is the "AVEVA Flex" subscription program, a system built around a proprietary currency known as Flex Credits. While pitched as a tool for agility and digital transformation, the model introduces a layer of financial abstraction that can be baffling for procurement teams and engineers alike.
Here is a breakdown of how Flex Credits work, why they are difficult to decipher, and who ultimately benefits from this complex business model.
At its simplest level, AVEVA Flex is a subscription intended to shift costs from Capital Expenditure (CapEx) to Operational Expenditure (OpEx). Think of it like an old-fashioned arcade. Instead of buying a specific game machine to keep in your basement forever (a perpetual license), you walk into AVEVA’s arcade and hand over cash in exchange for a bucket of tokens; these are Flex Credits.
The Purchase: You contract for a specific number of credits annually (e.g., 10,000 credits per year for three years). The cost per credit is negotiated, usually falling between $9 and $15 depending on volume.
The Spend (Burn Rate): Every piece of software you turn on "burns" credits at a specific rate. A PI Server might burn 20 credits a day. A PI Vision user might burn 1 credit a day.
The Flexibility: The appeal is that if you shut down a test server, you stop burning those tokens and can use them elsewhere; perhaps to spin up a cloud instance in AVEVA Connect.
The Complexity Trap: The theory is elegant; the reality is opaque. The difficulty in understanding your true costs stems from the fact that AVEVA has decoupled the price of the currency from the price of the product. To understand what a project will cost, you must now solve a two-variable equation where you don't fully control either variable.
- The Negotiated Rate: How much cash does one Flex Credit cost you? (Public records suggest around $13-$14.50 for mid-sized utilities).
- The "Rate Card": How many credits does the specific software configuration you need actually consume?
This second variable (the "burn rate") is the primary source of confusion. AVEVA does not publish a simple, universal menu saying "A 10k Tag PI Server = X Credits." Instead, these rates are hidden inside complex calculator tools managed by sales representatives. Consequently, an organization cannot simply budget for a project based on a price list. They must engage sales engineering to run a simulation to define how many "tokens" they need to buy to play the game they want to play.
Who Does This Favor? The Power of Obfuscation
When a vendor replaces a simple pricing model with a complex, currency-based system, it is rarely done for the customer's benefit. While Flex does offer genuine advantages for dynamic organizations that constantly spin up and tear down infrastructure, the complexity of the model heavily favors the vendor. And lets be honest, in the world of historian, how often do we spin up and tear down infrastructure?
1. Hiding Price Increases The most significant advantage for the vendor is the obfuscation of Total Cost of Ownership (TCO). A customer might see a lower upfront annual payment compared to a massive historical perpetual license purchase and think they are saving money. However, over a 5-to-10-year horizon, the subscription model almost always yields significantly higher revenue for the vendor. The credit system masks this long-term reality behind a layer of abstraction.
2. Vendor Leverage and Lock-in Because the "rate card" of how many credits a product consumes is not easily public, the vendor holds the knowledge advantage in negotiation. Furthermore, unlike perpetual licenses where the software keeps running even if you stop paying support, if you stop buying Flex Credits, the software turns off. This creates immense pressure to renew at whatever credit price is dictated at contract end.
3. Incentivized "Spend" Like gift cards, unused Flex Credits expire at the end of the term. This creates a "use it or lose it" mentality. Customers often find themselves with leftover credits and are encouraged by sales reps to "try out" new cloud services rather than let the credits vanish. This successfully embeds deeper parts of the AVEVA portfolio into the customer's operations, ensuring a higher credit requirement in the next contract cycle.
In summary, AVEVA Flex Credits are a sophisticated financial tool designed for the modern SaaS era. They offer technical flexibility, but they demand financial vigilance. When the price tag is replaced by a proprietary currency, the exchange rate is almost always set in favor of the house.
You probably are not surprised to learn that there is a drastic price savings between Flow Software and AVEVA, but does that alone warrant a change? PI’s licensing model shapes engineering decisions. Tag limits force teams to store less than they want. Modern analytics demand more complete history and deeper context.
AF is also tied to the PI historian. Most plants now use mixed sources. They have OEM systems, niche historians, cloud data, and edge IIoT signals. AF cannot cleanly span that landscape. When the modeling layer is bound to one backend, it cannot adapt as the plant evolves.
The architecture itself has aged. High performing operations split raw storage from contextual modeling. One system collects everything. Another defines meaning. PI blends those roles. That blend complicates growth and creates friction for AI and cloud workloads.
Timebase provides unrestricted historization. You record everything without licensing pressure. This creates several advantages. You build a complete archive. Engineers stop filtering data to manage cost. You can explore long-term trends and support modern AI workloads because the data exists.
You can log diagnostics freely. Investigations are faster because temporary or exploratory tags have no licensing impact. You gain a foundation designed for AI use cases. Timebase handles high throughput and long retention while staying simple to operate. Expansion becomes operational, not financial.
InfoHub replaces the AF concept and works across any time series source. It reads from PI, Timebase, and other historians. It becomes the home for equipment structure, KPIs, calculations, and events.
InfoHub handles:
• asset modeling and inheritance
• incorporation of data from SQL databases and other time series databases
• data cleansing, calculations, and KPI aggregations
• event detection and state logic
• data governance
• publishing for UNS or cloud analytics
This separation matters. If a tag moves from PI to Timebase, the model stays intact. If a site brings in data from a new vendor system, InfoHub can absorb it. You keep the organization AF delivered but remove the vendor lock.
Start by logging PI tags into Timebase in parallel. You keep PI in place but build a full, unrestricted archive.
Rebuild AF structures inside InfoHub and point them at PI to validate the model. Once results match, shift selected KPIs and events to Timebase inputs.
Publish from InfoHub so downstream consumers continue using the same datasets and endpoints. The transition stays invisible to them.
Retire PI when enough of the model runs on Timebase and InfoHub. You control the timeline.
InfoHub supports the core AF concepts. Asset hierarchies, calculations, inheritance, events, and reprocessing all carry forward. The difference is reach. InfoHub applies these rules across multiple time series sources, not only PI.
This lets you standardize KPIs, align reporting, and unify analytics across diverse systems.
You keep what AF accomplished. You remove the constraints that limit its future. Timebase becomes the unrestricted historical backbone. InfoHub becomes the modeling and governance layer that spans all your time series sources.
The plant gains a system that scales, adapts, and supports modern analytics without being tied to a single vendor’s licensing model.
The Time-in-State Metric (TISM) provides a structured approach to evaluating process stability and performance by measuring the duration a system remains in an operational state. By implementing TISM with Flow Software, organizations gain structured, high-confidence insights into process efficiency, variability, and improvement opportunities.
The need for robust Industrial Data Management (IDM) solutions has never been greater. Whether integrating operational technology (OT) with enterprise IT systems, generating the KPIs that drive the business, or supporting advanced analytics, manufacturers require a platform that is flexible, scalable, and delivers ROI in days not years.
Manufacturing leaders and plant managers often rely on metrics like OEE and APQ to gauge operational efficiency and identify areas for improvement. While these metrics are powerful, their effectiveness depends heavily on how they are applied. Incorrect calculation or over-reliance on OEE as a standalone KPI can lead to misleading insights.
At Flow Software, we exist to give teams decision-making superpowers.
We’re a driven, curious, and dedicated global team, committed to turning raw data into actionable insights since 2011.
Guided by our core values, we empower businesses to elevate their operations, embrace innovation, and achieve lasting success. Remote by design and lean by choice, we deliver solutions that simplify complexity and inspire smarter decisions.
When information flows, insights grow.