A machine learning enabled predictive maintenance solution which promises to be affordable for manufacturing SMEs will be developed through a partnership between Industry 4.0 start up SamsonVT and industrial products distributor RS Components.
The £1.3 million ADX project will focus on the application of anomaly detection for improved maintenance, engineering and decision making. It will see SamsonVT integrate data extraction, criticality assessment, machine learning and root cause analysis with its existing condition monitoring platform, SamsonBASE.
The end result is expected to be a predictive maintenance platform capable of using standard manufacturing environment equipment to harvest and process relevant data, in order to detect anomalies within machinery.
The partnership is backed by Innovate UK, which has awarded a joint R&D grant to support the technology’s development.
“We know that the high costs and complexity of PdM tools are a big barrier to adoption for SMEs, which make up the majority of the UK’s manufacturing companies,” said SamsonVT CEO Sam Burgess. “By delivering a PdM platform that is accessible and affordable – leveraging cutting edge machine learning techniques – we can help save British manufacturers billions every year in unplanned machine downtime.”
Machinery downtime is known to cost manufacturers in Britain around £180 billion every year, with some large manufacturers losing over half a million pounds annually.
Implementation of a predictive maintenance solution could help SMEs save 65% of their downtime costs as well as enabling them to optimise maintenance resources, reduce the unnecessary replacement of parts – as often seen with time based maintenance practices – and minimise the ancillary damage repair costs caused when parts fail in situ.
However the majority of SMEs are yet to adopt a predictive maintenance approach due to high implementation costs. The ADX project aims to provide them with bespoke machine learning / anomaly detection models for effective maintenance, with no initial costly changes to infrastructure required.
This means manufacturers can use their existing network of sensors, hardware, and equipment – provided they have been harnessing data for a minimum of 3 to 6 months – and do not require highly trained personnel to interpret the findings.
RS Components’ director of maintenance solutions Richard Jeffers said: “We are excited to be working with SamsonVT on this project. For busy SMEs that may not have previously considered PdM as an option, due to perceived cost and limited management time, this will be a real gamechanger. This will provide them with a PdM solution that is as close to plug and play as you are going to get, generating the insights they need in order to know when to act and, just as importantly, when not to.”