Industrial AI
Optimize your production process using cutting edge technology
We at Zyfra optimize industrial processes
using Deep Learning and numerical
optimization methods
By applying cutting edge data science combined with industrial expertise, we
provide end-to-end solutions for optimizing production speed, quality, and costs.
Metals (steel and aluminium production)
Oil & Gas
Increase production and reduce costs
Mining industry extracts valuable materials from the earth and supplies commodities for the steel and other metals industries, energy industry and even high-tech-industries (rare earths).
Mining process requires a lot of capital and human resources, and impacts both environmental and human safety.

We in Zyfra utilze latest developments in AI in order to improve mining safety and operational efficiency, reduce mining costs. Our mission in mining is to drastically decrease a need for human presence at the mine, making it virtually humanless.
Increase production and reduce costs
Steel is being produced from iron ore and scrap in blast furnaces or electric arc furnaces (EAF). Then molten steel is being further processed in ladle furnaces. After that the steel is being casted and hot-rolled. Energy constitutes a significant portion of the cost of steel
production, from 20% to 40%.

AI could optimize a whole EAF steelmaking process, improving energy and ferroalloys consumption and increasing a product output and effectiveness of the production process.

Machine Learning methods are also capable to increase an efficiency of steel products making - including pipes, rails, wire, etc. From our experience, application of AI could increase throughput of steel products heat treatment process by 5% and more.

Computer vision and machine learning method are also prospective for aluminium production efficiency improvement. Production of aluminium consumes vast amounts of energy, more than 3% of the world's entire electrical supply, which exceeds a percentage of energy consumed by the steel producton.
Oil & gas industry performs exploration, extraction, refining and transportation of oil & gas and their products. Oil & gas industry is vital for the civilization, but requires enormous efforts and utlilization of the latest technologies to be able to support and increase production levels.

AI will have a huge technological impact on the oil and gas industry over the coming years. Zyfra has a track record of application of Machine Learning to some of most challenging problems in O&G, including:
  • Drilling anomalies prediction and prevention
  • Oil extraction and reservoir pressure support optimization scaled up to the thousands of wells
  • Petrochemical production processes optimization
  • Petroleum products demand prediction
  • Oil refinery plant logistics and production schedule optimization.
The chemical industry is central to the modern world economy, it converts raw materials (oil, natural gas, air, water, metals, and minerals) into tens of thousands different products.
Chemical products include polymers and plastics, bulk petrochemicals and intermediates, other derivatives, inorganic chemicals, and fertilizers. Life sciences chemicals, like pharmaceuticals, health products and pesticides, and specialty chemicals like electronic chemicals, adhesives, sealants, coatings and cleaning are products of the chemical industry too.

Application of AI, high-performance computing and quantum computing could revolutionize a lot of aspects of chemical industry, starting from the automation of new materials and biological design processes, to the manufacturing optimization, including virtual sensors to measure chemical production process quality, real-time process optimiziers (RTO), to predictive maintenance and industrial safety, to the demand forecasting and logistics and business process optimization.
AI Heat Treatment
AI Smelting
Increase steel products heat treatment throughput by more than 5%
Heat treatment is being used to alter physical properties of steel products, like pipes, rails and wire. Heat treatment involves the use of heating and chilling, normally to extreme temperatures, to achieve a desired result such as hardening or softening of a material. Heat treatment may include quenching, tempering and normalizing.

Zyfra solution for Heat Treatment Optimization allows to increase the productivity of the heat treatment shop while mantaining the quality of end product. It delivers real-time recommendations on the optimal parameters of the heat treatment process (temperatures, speed etc.).
Solution consists of two major modules: prediction module, which predicts expected quality using source data like steel grade, pipes diameter and thickness, temperatures and speed, etc. and optimization, which uses generated prediction model and optimization methods in order to provide recommendations on best available treatment regime.

Heat treatment process throughput is expected to increase by more than 5% as a result of the process optimization.
Produce more steel
Steel production using Electric Arc Furnace (EAF) is responsible for 28% of world steel production and this figure is expected to rise as soon as more and more metal scrap will be available. EAF process can consist of following stages:
  • EAF
  • Ladle Treatment
  • Vacuum Treatment
  • Continuous Casting.

Our solution zSmelting allows to analyze data collected during steelmaking process and optimize its parameters on all stages, providing following benefits:
  • Reduction of energy consumption
  • Reduction of ferroalloys consumption
  • Increase of the product output
  • Decrease of the production scrap rate
  • Decrease of the unit cost
  • Increased total effectiveness of the production process.
    Optimize plant operation for the best margin
    Scheduling gets the more complicated the more factors you take into account. For decades we had to let go some factors or conditions to achieve reasanable performance.

    Thanks to advances in AI, we now can build schedules much faster while considering many more factors.

    Additional perk is the AI's ability to learn restrictions that were present but not recorded in any data as is.

    Improving scheduling with our solution provides up to 10% increase in margin.
    Using AI to reduce logistics costs by 15%
    Both predicting demand and optimising logistics are hard tasks, and AI allows to achieve better results compared to classical optimization models such as linear programming.
    Additional perk is the AI's ability to learn restrictions that were present but not recorded in any data as is.
    15% cost reduction is the calculated result of our project in a large oil company.
    Using AI to reduce electricity consumtion by 20%
    Chillers are used to keep optimal temperature inside the plant. They are controlled by humans, who can change settings several times a day, but conditions. We use AI-powered system that automatically selects optimal modes while keeping acceptable temperatures.

    The system gives recommendations to switch each chiller on or off, as well as optimal temperature settings.

    The result of applying such a solution is up to 20% electricity consumption reduction.
    Zyfra Eye
    Virtual Sensor
    Auto Analog
    AI Platform
    Zyfra Eye is a Computer Vision and Deep Learning powered solution which processes raw video footage of the dashboards of non-connected industrial machines and other industrial equipment, recognizes meters readings and controls positions, converts them to the structured data and transfers to SCADA or IoT platform.

    It comes as a server software for video feed processing or as a hardware solution, combining industrial camera and video recognition unit.

    Zyfra Eye has a convenient UI, walking you through the installation and setup process.

    Zyfra Eye allows you to monitor your unconnected industrial equipment for a fraction of price of the traditional machine data collection installation.
    Measure things deep inside technological process
    Zyfra vSensor (virtual sensor) allows to identify quality metrics of chemical industry products and semi-products in real time when it's impossible or very expensive to conduct direct quality measurements (e.g. in aggressive and hazardous environments).

    By using mathematical model based on the measurable metrics and results of precision quality control in a lab environment, vSensor allows to obtain near real-time quality control. It allows to save more than 90% of the price of the gas chromatograph.

    Finding similar things despite
    different SKU
    Manage ML/AI services in a single view
    The platform behind our solutions and products provides data integration, validation, model training, applying and scoring, API, monitoring, GUI as an enterprise-grade system ready to provide value by allowing fast movement from ad-hoc data ideas to production service.

    The GUI provides visibility into optimization effects, controlling how prognouses and recommendations were used and other useful insights into AI applications in your infrastructure.
    Digital Adivsor & Predictor
    We can use your data and produce value for any process where cost, value or quality worth a lot.

    Our solutions can either predict quality and product properties, or recommend process settings and material usage.
    Big Data & Data Lake
    Our huge expirience building DWH and Hadoop Data Lakes can be utilized for your benefit, providing enterprise integrations, BI and security.
    Audit & Consulting
    We provide services auditing your company's data handling, identifing good opportunities to implement AI solutions and translate possible effects into USD value.
    Learn more about our solutions, products and services
    We will get back to you ASAP. Real industry talk, no crap.
    Ask us whatever you're interested in. Our AI is capable of handling any questions.
    By clicking 'Contact Us' button you agree with our Privacy Policy

    Html code will be here

    Our contacts
    Feel free to write and call us. We really love to communicate with our clients.

    +358 (0) 9 427 34852

    Pasilankatu 2, 00240 Helsinki