Making Factories Smarter Through Machine Learning
2. BACKGROUND
Plethora IIoT 10, part of the ETXE-TAR Group, a federation of closely coupled industrial manufacturing companies, through use of machine learning is utilizing predictive maintenance to minimize downtime and maximize the operation of their sophisticated Computer Numeric Control( CNC) machines. These machines are used to build automobile powertrain parts, like crankshafts, camshafts and connecting rods, with highly precise requirements executed over an extremely short cycle time. A key advantage that Plethora utilizes is monitoring key parameters to effectively learn the system behavior to predict failures before they occur. With the support of ETXE-TAR and Ikergune, the research and development company for the ETXE-TAR Group and other ecosystem partners, Plethora IIoT has developed a system called Oberon. This system covers the needs from data gathering independently from the domain to visualization, giving important information in terms of machine behavior( useful for machine or system builder) and process behavior( useful for machine / system user).
One of the key elements of the Oberon system is the intelligent gateway, designed and manufactured by System-on-Chip Engineering( SoC-e) 11. The gateway utilizes highly configurable all programmable System-on-Chip( SoC) technology to perform real time acquisition, sensor fusion, filtering, pattern detection and analysis of the data extracted. The gateway is intelligent in the way that it combines smart networking, flexible hardware-software computation and high speed data acquisition systems. This system is able to provide“ Plug & Work” operation through time-sensitive, redundant protocols for Ethernet( High Availability-Seamless Redundancy / Parallel Redundancy Protocol- HSR / PRP). These protocols ensure zero loss of data in the event of network failure or hot plug-in due to plant layout changes. From the security and cybersecurity point-of-view, the intelligent gateway ensures a multi- layered security approach. Strong security means are applied at the device, system, network and application levels. This approach, based on cybersecurity by design is a key element in the emerging connected industrial environment. Learnings from the Industrial Internet Consortium’ s( IIC) Industrial Internet Security Framework( IISF) document are directly applied, as this is the same gateway found at the heart of the IIC Security Claims Evaluation Testbed. Details of the intelligent gateway functionality and capabilities applied to the Smart Factory are presented in XCell Journal 92 Article“ Intelligent Gateways Make the Factory Smarter.”
In terms of acquisition, sensor fusion, pre-processing and processing software developed by Ikergune, the main objective is to gather data coming from sensors with the required sampling rate, independently from the variable. Therefore, it can sample each variable with smart criterions: For example, temperature may not be measured with the same frequency of vibration
10
SPRI, Basque Industrie 4.0, 29 Nov 2016. [ En línea ]. Available: https:// goo. gl / jrzFlF
11
System-On-Chip engineering – SOC-e CPPS 40 Intelligent Gateway.
IIC Journal of Innovation- 31-