Novel software monitors manufacturing, saves millions
School spin-out company, Perceptive Engineering, markets software to monitor and control manufacturing processes. Its technology reduces the manufacturing costs of customers and also allows them to increase productivity without the need for expensive upgrades to process plants. The company now employs more than 25 staff, has won several awards and in 2013, opened its first overseas office in Singapore.
Manufacturing and processing industries have been slow to adopt statistical techniques for detecting faults in equipment and plant. The complexity of this approach made it virtually impossible to implement in real-time fault monitoring.
In response to requests from industry, School researchers developed new, more robust algorithms that finally brought the power of multivariate statistical analysis to real-time, industrial process control and condition monitoring.
The technology improves plant efficiency, reduces operating costs, increases productivity and reduces energy use in a wide range of industrial sectors. Since 2008, Perceptive has reached the finals of The Engineer Innovation Awards and the UK IET Innovation Awards.
Today Perceptive employs more than 25 staff and is actively recruiting more. Many of the company’s employees are MSc and PhD graduates from the University of Manchester.
Perceptive systems are used by manufacturers and processors in most industry sectors, from global chemicals and steel manufacturing to food and utilities companies. Examples of efficiency gains and cost savings in a selection of sectors include:
- Pharmaceuticals and nutritional powders
Perceptive Engineering’s control systems have been applied to plants in Europe and the USA, providing considerable improvements to process yield and increased operating capacity and overall efficiency.
Perceptive Engineering work with a number of major UK water companies, where their control systems have improved plant efficiency considerably. These control systems have also decreased CO2 emissions and resulted in payback periods of several months. There are plans for the wide-scale adoption of Perceptive’s monitoring and control technology across wastewater treatment plants in the UK.
The companies control systems are in use in ten pulp and paper plants in Europe, Australia, Asia and the USA. These systems are used to reduce steam usage and cut CO2 emissions, which has resulted in significant financial savings.
Researchers from the School developed algorithms that use multivariate statistical techniques to analyse condition and production monitoring data. This statistical modelling can identify – and even predict – faults and inefficiencies in industrial and manufacturing processes in real-time, enabling much better automated control within industrial plant.
- Multivariate statistical algorithms that detect faults in real-time.
- Robust algorithms for integration into plant safety systems, able to shut down equipment when necessary.
- Self-monitoring controller technologies, able to remodel or retune their performance as required.
- Statistical techniques that enable controllers to use poor-quality sensor measurements.
- Series of algorithms for simple, accurate monitoring and control of batch processes.