Powder processing plants uptime increased
Hosokawa Micron uses XpertFactory to develop its Gen4 remote monitoring solution. The solution supports condition monitoring for preventative maintenance, on-line diagnostics for improved plant and product quality and trouble-shooting bots for minimising downtime on its milling equipment. This has resulted in 15 percent improvement in uptime; 20% reduction in energy usage and capacity gains of about 10% being recorded by Hosokawa customers.
Reduction in energy use
Reducing CO2 emissions at an oil company
Machine learning was used to understand how energy costs and emissions were dependent on external factors, control settings and operating practices. We identified opportunities for reducing energy costs by 3% and reducing CO2 emissions by 7%. Models of critical performance factors were also developed for use as benchmark for intelligent monitoring.
Reduction in CO2 emissions
Reducing energy costs at a UK brewery
Machine learning was used on data from a brewery central refrigeration system to understand the variations in the efficiency of the plant. The machine learning models identified opportunities to reduce energy costs by changing the control strategy and operating practices with minimal capital cost.
Annual reduction in energy costs
Reducing cost at a chemical complex
The objective was to identify measures to reduce operating costs. Machine learning patterns were identified relating the overall revenue generated to the operating parameters and modes of operation. These patterns were used to determine the optimal control set points and operating strategies required to minimize costs which resulted in a 5% reduction in costs.
Reduction in costs
Machine learning was used to understand and improve the performance of Alumina extraction from Bauxite. The model identified potential for a significant increase in performance from digestion estimated at $650K pa. This performance increase can be achieved using optimal control settings delivered as advise to operators and engineers.
Reduction in costs per annum
Reducing Nox emissions
We carried out a machine learning analysis to understand the operating patterns determining the levels of NOx emissions. The patterns we discovered related emissions to the type of coal being used, the use of other fuels such as gas and biomass, the required unit generating load, and the control settings on the unit such as damper positions, air flow rates and excess oxygen as well as which mills are in service. The discovered patterns were deployed as part of an on-line monitoring solution to improve the performance of the station by advising the operator on the optimal control settings for a given set of operating parameters.
Reduction in Nox emissions
Machine learning was used to analyse the performance of rotary kiln driers in an ore refinery to identify the best operating conditions to minimise energy costs while still meeting production and quality targets. Resulting machine learning model was used for online monitoring and advisory systems to optimise drier operating conditions saving 25% of energy use.
Savings of energy use
Want to find out more?