Machine vision uses cameras, computers and software algorithms for carrying out inspection tasks that require precise and repetitive verification and testing in high speed. Accomplishing such tasks with human vision is extremely difficult, if not impossible, because while human eyes are capable of making precise measures in details, they aren't equipped to do so in a rapid and repetitive manner; human vision, by comparison, is error-prone and could deteriorate irreversibly when made to perform such visually intensive tasks.

Introduction

Machine vision uses cameras, computers and software algorithms for carrying out inspection tasks that require precise and repetitive verification and testing in high speed. Accomplishing such tasks with human vision is extremely difficult, if not impossible, because while human eyes are capable of making precise measures in details, they arent equipped to do so in a rapid and repetitive manner; human vision, by comparison, is error-prone and could deteriorate irreversibly when made to perform such visually intensive tasks.

Replacing human vision with machine vision for verification and testing tasks is a part of industrial automation, intended for achieving faster TTM and optimizing resources while at the same time lowering cost of ownership and improving risk management.

Successful implementation of machine vision for all production processes requires the upmost care in not only software algorithms compilation and system hardware selection, but also the software and hardware integration.

Customer Requirements

A global leader who specializes in smart, safe and sustainable machine vision technology learnt about Lanner while seeking hardware solutions for building a factory automation system for inspecting recycled glass bottles at beverage packaging factories.

This packaging inspection system would automate the most stringent and thorough check on all recycled bottles for any kind of defect, determining their reusability. The top three most imperative requirements for this particular systems hardware are first of all, a high-performance CPU; second of all, a fast GbE interface; and third of all, rugged encasing. The CPU must be powerful enough for enabling glass inspection at a rate of up to 42,000 bottles per hour while the I/O interface must be flexible enough for supporting the installation of multiple cameras and the connection to the factory network for remote management. Finally, the hardware must be reliable so that it is available 24/7, requiring the absolute minimal amount of manual maintenance and troubleshooting.

Lanner Solution

Lanners LEC-7070, a fanless box PC powered by Intels Core i Series processors, was eventually selected, among other suppliers offerings, as the most suitable hardware on which the bottle inspection system could be built.

This fanless box PC is based on Intels Ivy Bridge microarchitecture and offers a rich I/O interface. It strikes the perfect balance between performance, cost and combines processing power, industrial-grade I/O functions, image-capturing and network communication capabilities in a compact form factor.

Customer Benefits

The LEC-7070 comes equipped with either the Intel Core i7-3517UE or the Core i3-3217UE processor, coupled with the Intel HM65 chipset. While both variants offer dual-core, quad-thread processing power and Intels Hyper-Threading Technology, the low power, generation three Core i7 variant also features Intels Turbo Boost Technology. This combination of powerful CPU, built-in proprietary technologies from Intel and up to 4GB of 1333MHz DDR3 system memory makes both variants more than capable of providing optimized operating speeds and higher throughputs for tackling even the most demanding multi-camera vision applications.

Multi-camera setup on this bottle inspection system is made possible via the LEC-7070s dual GbE LAN ports, enabling multi-vision for inspecting different parts of a bottle from different angles simultaneously. The VGA and HDMI video-out ports, on the other hand, deliver crucial images or video footage to multiple displays at the managing location for decision making and prompt actions. More specifically, each of these cameras carries out a specific task, varying from scanning and checking bottle surface for defects, foreign matters and other irregularities from top to bottom, inside and out, to capturing and delivering high definitions images/videos, via the LEC-7070s dual 10/100/1000Mbps autosensing LAN ports, to multiple displays for determining a bottles suitability for reuse.

Whats more, the LEC-7070 can be configured, via the SIM card reader built into one of the mini-PCIe buses for wireless/3G connections. In addition the aforementioned I/O interface, the LEC-7070 also provides 4 USB ports, 1 DIO port and two COM ports for simplifying hardware integration.

Last but not least, the LEC-7070s aluminum chassis and fanless design make it completely dustproof, reducing downtime and maintenance overhead. Whats more, when outfitted with industrial-grade components, the LEC-7070 is operable under extreme temperatures ranging from minus 20°C to 50°C, making it the ideal hardware solution for multi-camera machine vision applications found in most factory floors and production settings.

Conclusion

Having successfully using Lanners LEC-7070 for bottle inspection, Lanners hardware solutions are currently being considered by this particular system integrator for the development of other automated multi-cam inspection systems.

These systems will be deployed not only for inspecting fill level, product labels, character recognition and container molding in the food and beverage industry but also for other multi-vision verification and testing tasks at vehicle manufacturing facilities.