How Machine Vision, Optical Inspection and Advanced Imaging Is Improving Industrial Efficiency
Jun 26, 2021
Advanced sensors, AI, machine learning and IIoT systems represent the greatest opportunities for efficiency, reliability and productivity gains for many industrial businesses.
Developing tools that help businesses do more with limited resource, improve product consistency, eliminate waste and perform more reliable quality assurance at each step of the production process isn’t just good for manufacturers, it’s good for every consumer.
Industrial Edge Computing and Deterministic Processing
Many of the industrial computing technologies developed by SMART Embedded Computing are designed to:
Enable and enhance industrial automation
Improve process control
Increase data acquisition capacity
We’ve built upon our experience with industrial computing technologies to push the “edges” of edge computing. Our team is always working on ways to improve data aggregation and analysis, as well as aid in the development of IIoT and connected manufacturing platforms that make better use of machine learning and AI.
In What Types of Applications Are VME Single Board Computers Used?
VME single board computers offer exceptional value thanks in part to their long lifecycles. They often function as embedded control platforms for:
Semiconductor processing equipment
Medical diagnostic and imaging technologies
Various medical therapy devices
The multiprocessing bus capabilities in these single board computers allow customers to get more functionality and power out of these small but technically robust devices.
Many industrial facilities rely on connected technologies, but limited bandwidth and physical distance can result in latency, reduced performance and potential instability. As industry 4.0 technologies are more widely adopted, edge servers offer an on-site, facility-based computing solution that minimizes bandwidth and latency problems. They are often vital pieces of equipment for a business that is leveraging inference engines, machine learning or artificial intelligence-powered vision systems that perform:
Advanced image processing
These types of machine learning and AI-powered processes are reducing the costs of maintenance, reducing frequency of breakdowns, improving efficiency and making defects fewer and further between.
Examples of Machine Learning and Machine Vision in Industrial Applications
In nearly every industry where a product is being manufactured and packaged for a customer that may be several distribution steps removed from the source, the reliability of packaging is hugely important. Pharmaceutical companies especially must hold themselves to the highest standards of packaging reliability.
For example, one pharmaceutical packaging company has developed image processing to:
Check for broken or malformed tablets
Analyze tablets to make sure they’re the right length, width and color
Make sure that the right number of capsules are in each bottle
Mistakes are logged and reported automatically for improved accountability and quality assurance in the production process.
Machine vision is also becoming increasingly important for particularly complex inspection tasks that humans frequently struggle to reliably accomplish. Consider automobile construction. Electrical mistakes that are made early in the process can result in catastrophic problems later down the line. Cameras that can take many high-resolution images at different angles to develop 3D component models, including their connections, enable fast and accurate quality assurance inspections.
The Eyes of Automation
Those cameras can also be used in fast-moving consumer goods (FMCG) or food manufacturing, where cameras facilitate the entirely automated identification, sorting and stacking of food pallets. Some cameras can read bar codes for proper product classification and conveyer belt assignment while additional cameras can track the number of pallets, product types and even expiration dates going into each pallet. All of this data informs automated decisions regarding transport labeling and shipping.
Machine vision is already a vital tool in tracking highly regulated products like pharmaceuticals as they travel around the globe. Batch numbers, manufacturing dates, serial numbers and expiration dates can be associated with specific bottles of medication and stored in master databases as Global Trade Item Numbers (GTIN). Machine vision on the receiving end of those bottles can set aside or reject any packages or cartons with GTIN that don’t match the master database or those that may be flagged for recall or disposal.
When critical industrial equipment breaks down unexpectedly, companies can suffer sometimes serious delays or even production shutdowns. IIoT devices that act as the sensory organs for machine learning can monitor all types of machinery, component variables and data to identify potential problems before they happen and automatically suggest or take corrective maintenance action.
When minutes of downtime can result in tens of thousands of dollars of lost revenue in some industries, there’s almost no investment in predictive maintenance that’s too great.
Leverage Your Own Machine Vision, Optical Inspection and Predictive Maintenance Solutions
SMART Embedded Computing develops edge server technologies that facilitate the development and smooth functioning of an array of IIoT, machine learning and AI systems.
You can learn more about SMART EC’s extreme edge servers by visiting our website, reading our whitepapers or contacting us at 602-438-5720.