Factory Harmonizer
icon_soft-sensors_white

Soft sensors

digital colour visual

Soft sensors, also known as virtual sensors, provide real-time estimates of variables that are difficult, costly, or even impossible to measure directly in a production environment. By leveraging advanced algorithms and machine learning techniques, our soft sensors use available process data to infer continuous estimates for critical parameters.

AI green button

Why choose soft sensors?

Understanding every aspect of your production process is key to staying ahead. Soft sensors enable you to:

  • Gain Real-Time Insights: Get continuous visibility into your production parameters, allowing for immediate adjustments and optimizations.

  • Reduce Costs: Minimize the need for expensive testing and reduce unnecessary consumption with the gained information. 

  • Improve Process Efficiency: Optimize your production process by accurately estimating key variables, leading to improved quality, reduced waste, and lower energy consumption.
Green technology gear on a motherboard

How soft sensors work

Soft sensors are built by combining process knowledge with advanced data analytics. Here’s how we do it:

We start by gathering historical process data from various sources, such as Distributed Control Systems (DCS), Manufacturing Execution Systems (MES), and laboratory data.

Collected data is cleaned, normalized, and filtered to remove any noise or inconsistencies, ensuring high-quality inputs for model development.

Using machine learning algorithms, we create models that predict the values of difficult-to-measure variables based on the available data. These models are rigorously tested to ensure accuracy and reliability.

Once the models are refined, they undergo thorough validation to ensure they perform as expected in real-world conditions. After validation, the models are deployed into your production environment, where they continuously provide real-time estimates.

Use case example

Chemical Production

Soft sensors enable real-time estimation of critical quality parameters such as pH levels, temperatures, pressures, and raw material and energy consumptions. This allows for more efficient production processes, reducing waste and operational costs, and improving overall product quality and yield.

use case example

Pulp and Paper

Soft sensors provide precise monitoring of e.g. moisture content and other key factors critical to the papermaking process. By enabling tighter control over these variables, they help maintain consistent product quality, minimize material waste, and improve energy efficiency across production stages.

Use case example

Metal Industry

Soft sensors offer real-time monitoring of essential metal refining and production parameters such as flow rates of raw materials, catalysts and utilities, temperatures, pressures, concentration and ratios of involved material flows, impurity levels, reaction times and reaction balances. This allows operators to adjust treatment processes proactively, ensuring compliance with environmental regulations and optimizing the efficiency of metal refining processes.
Blue light swirl

Overcoming Common Challenges

While the benefits of soft sensors are clear, successful implementation requires careful planning and execution. At SimAnalytics, we address the challenges of soft sensor design by:

  • Ensuring Data Quality: We perform a comprehensive Data Health Check before deployment to ensure only relevant and high-quality data is used in the models.

  • Customizing Models: Each soft sensor is tailored to your specific process needs, ensuring that the models provide the most accurate and useful insights. We ensure a good dialogue with your process experts throughout the development process.

  • Seamless Integration: Our soft sensors use our advanced Factory Harmonizer Data Platform to smoothly integrate with your existing systems. This also enables updates without disturbances to your production. 

Our Experts are here to help!

Johanna Kummala
Head of Sales

Book a demo