As an Information and Communications Technology (ICT) leader with 120 years of expertise, NEC has leveraged its adaptive solutions to enable individuals and enterprises across a wide spectrum of industries to achieve more flexible, secure and responsive work environments.
As an Information and Communications Technology (ICT) leader for 120 years, NEC enables new approaches to how Communications and IT services are delivered and managed through its field-tested expertise.
Combining innovative capability with a rich Communications and IT portfolio, NEC has provided individuals and enterprises with solutions that cover and enable the full spectrum of business processes for new and improved ways of conducting business.
Deep learning that automates factory defect inspection
When quality control fails the consequences can be damaging.
Potential fallout can include high-profile recalls, dissatisfied customers, customer loss and a tarnished reputation.
Automated imaging-based analysis using machine vision systems offer manufacturers a certain level of assurance in defect inspection as do
manual reviews by trained and qualified inspectors. Yet these safeguards can be inconsistent with some anomalies too difficult to catch
making quality failure possible.
Our Alternative Solution
NEC brings greater control to manufacturing defect inspection through its AI technology.
Automated visual inspection using NEC RAPID Machine Learning - the company's premier deep learning software - cuts inspection time
by recognizing a wider variety of defects than conventional machine vision techniques. NEC RAPID Machine Learning also improves quality control consistency
by reducing the need for manual inspections to lessen the possibility for human error. Additionally being able to detect defects on diverse
surfaces, NEC RAPID Machine Learning is highly adaptive making it an ideal solution for a range of manufacturing industries.
From possible applications and proofs of concept to an actual automotive case study, this white paper explores NEC's leadership in industrial visual
inspection using deep learning IA.
Feed image data will be used by NEC RAPID Machine Learning to create its defect inspection engine
which can be utilized by users to inspect defects automatically.
Image analysis requires a huge number of pictures of both good and defective products to be collected for learning by AI,
but collecting images of defective products is no easy matter in the precision manufacturing industry.
AI-based surface inspection algorithm called "One Class" is a new feature of RAPID machine learning that is extremely powerful
for manufacturing inspections. Manufacturers do not need to collect images of defective products.
The "One Class" Advantage
This video demonstrates how "One Class" enables defective products to be identified just by learning images of good products.
NEC RAPID Machine Learning's Core Deliverables
Ready access to reliable insight based on real-time information.
The ability to make meaningful changes that deliver both time and cost savings.
The capability to identify potential issues before they become detrimental situations.