Property Estimation
When You Can't Measure It Directly, Estimate It Accurately
Some of the most important quality and process variables in your operation aren't things you can measure continuously. Lab-tested properties, product composition, and certain physical characteristics often depend on infrequent sampling — sometimes hours apart — leaving operators making decisions on outdated information between tests. Capstone Technology Canada's Property Estimation solutions close that gap using soft sensor technology: virtual measurements that give you a continuous, real-time read on properties that would otherwise require waiting on a lab result.
How Soft Sensors Work
Capstone's Estimator uses readily available, continuously measured process variables — temperature, pressure, flow, and similar signals your plant is already collecting — to infer values for properties that are difficult, expensive, or infrequent to measure directly. Rather than installing additional physical instrumentation or waiting on the next lab sample, the Estimator models the relationship between what you can measure constantly and what you actually need to know.
Our technology supports multiple modeling approaches, including time series analysis, state space models, Kalman filters, and neural networks, and automatically selects the most reliable method for a given operating regime. That flexibility matters because no single modeling approach performs best across every process condition — a technique that works well during steady-state operation may lose accuracy during a transition or upset, so the system adapts rather than relying on a single fixed model.
What Property Estimation Solves
- Continuous quality visibility — See an estimated property value updating in real time, instead of only at the interval your lab samples allow
- Faster response to quality drift — Catch a deviation minutes after it starts, not hours later when the next lab result comes back
- Reduced dependence on physical sensors — For properties that are costly or difficult to measure with hardware, a soft sensor delivers a reliable virtual measurement without additional capital investment
- Better decision-making between lab samples — Operators aren't left guessing what's happening to product quality in the gaps between tests
Beyond Quality: Cost and Profitability Insight
Property estimation isn't limited to physical or chemical properties — the same approach applies to evaluating operational costs across your process. By analyzing the activities and variables that drive cost in specific sections of a plant, this same estimation approach can surface improvement opportunities that wouldn't be visible from standard reporting alone — for example, tools that show how profitable a specific unit or process area is in real time, which can help justify targeted maintenance or capital investment decisions with clear, current data rather than a retrospective cost report.
Proven in the Field
Capstone has applied soft sensor and property estimation technology across a range of industrial applications — from estimating pellet physical properties in mineral processing to inferring product quality in refining and chemical operations. In one mining implementation, automatic estimation of pellet physical properties was a key feature of an advanced control solution that contributed to a measurable increase in production performance.
Why This Matters for Your Plant
If your team is making quality or cost decisions based on data that's hours old by the time it arrives, property estimation gives you the same insight in real time. Whether the goal is tighter quality control, reduced lab dependency, or clearer visibility into where cost is actually being generated in your process, soft sensor-based estimation turns data you already have into information you can act on immediately.
Talk to a process control engineer about property estimation