Unexpected equipment failure is a serious issue for businesses, stopping a line for an undetermined amount of time and leaving products unfinished while technicians implement repairs or replacement. These setbacks are costly and always undesirable.
Using machine learning, however, it’s possible to detect that a piece of equipment is starting to approach failure well in advance of the problem being noticeable through typical human observations. This applies especially to subtle changes over time, where a normal threshold value approach would fail to pick up the deviation. One example is by sensing subtle shifts in vibration e.g. due to motor torque, embedded devices can detect when a machine is falling out of spec and notify a manager of its condition. This allows the falling parts of the equipment to be specifically maintained, rather than targeting multiple components on a set schedule (including potentially still functional ones), or worse, doing nothing and letting a machine come to a crashing halt.
The machine learning approach is feasible today using Edge Impulse’s trusted edge ML platform to train an embedded device to detect the anomalies that indicate an impending failure. The result: running longer and more efficiently, avoiding downtime while increasing profitability.
To help familiarize users with what the possibilities that predictive maintenance offers, we’re co-hosting a webinar with RealPars, the leader in technical training, on October 26, at 3pm GMT (8am PT). The free event will offer talks from engineers and executives from both companies, and focus on how predictive maintenance can help improve your business results and increase your ROI.
Event details
How Machine Learning Can Help With Predictive Maintenance for Industrial Applications
Date: October 26, 2022
Time: 8:00am PT
Predictive maintenance is a data- and technology-based method of carrying out proactive maintenance. The main goal of predictive maintenance is to identify potential machine issues well before they lead to critical situations with shutdowns. This is realized through constant equipment performance monitoring enabled by sensors, data collection, and near-real-time communication between equipment and software. The new rise of interesting automation equipment & parts enable this in part or can at least run C++, which is of great help in such a context. Predictive maintenance can offer numerous, unparalleled benefits in productivity and efficiency — benefits which you will be able to understand from this predictive maintenance session between RealPars and Edge Impulse. It is also important to understand predictive maintenance challenges exist, but are well worth the future ROI of an edge ML enabled predictive maintenance solution.
Speakers
Shahpour Shapournia
Co-Founder & CEO · RealPars
Shahpour has 8 years of experience as a PLC programmer mainly in the oil, gas, and steel industry. He understands the challenges of becoming a successful automation engineer, because he was once like his students — ready to grow, but not knowing how to go about it. He loves taking complicated engineering concepts and explaining them in an easy-to-follow format.
Ken Bourke
Senior Automation & Controls Engineer · RealPars
Ken Bourke has worked in the automation industry for close to a decade. Throughout his career, he has worked on global projects as a commissioning engineer, controls project engineer, controls architect, and software consultant. Since 2022, Ken has worked with RealPars in his free time developing training programs to make careers in automation more accessible to people around the world.
Mihajlo Raljic
Director, Sales & Business Development EMEA · Edge Impulse
Mihajlo Raljic has history of industrial engagements where he worked for over 10 years for companies such as BASF, Bayer, Boehringer Ingelheim, and Q-Cells, in roles including Production & Technology, Marketing, and Sales. Prior to joining the edge machine learning endeavors in 2019 he worked for another 10 years in renewable industry. Here he got a lot of exposure to solar photovoltaic production, technology, and the quirks of getting solar farms of the ground including all aspects such as grid connection/stability, permitting, and technical design including O&M for energy storage-based systems.
Louis Moreau
Senior DevRel Engineer · Edge Impulse
Louis started his career developing a low-power and GPS-based IoT solution to protect rhinos in an African conservancy. He is now working at Edge Impulse as Senior DevRel Engineer. When not hacking development boards or automating repetitive tasks, you can see him riding an electric skateboard in Lille, northern France, or scuba diving in some paradise places.