ReliVision is hardware and image-modality agnostic and applicable across different use cases,
enabling quality automation and smart manufacturing without writing a single line of code.
shop floor availability
ReliVision uses machine learning to achieve a high level of flexibility and scalability for computer vision systems.
Select an application
As part of its industry 4.0 transformation, our customer is implementing Relimetrics' AI accelerated machine vision technology at Foxconn to automate quality inspection of server assembly.
Our customer's portfolio of server products is highly complex and customized. The assembly of different server types must be inspected in-line near real time. This problem is well suited for Relimetrics' AI-based machine vision software.
For more details visit our insight.tech article.
Increases detection accuracy
Reduces assembly audit time
30-40 seconds per assembly audit
Reduces final testing and rework
"Relimetrics solution is embedded within our assembly line and enables immediate feedback to the operators and line managers to address any occurring issues right away. The fact that Reli-QA has no fatigue-induced errors and handles bespoke tolerances drove a 2% point increase in the initial pass rate. This is significant as it moved the performance from 2.1σ to 4.2σ."
Lockheed Martin is working together with Relimetrics to automate aircraft manufacturing and assembly processes with Relimetrics' AI-based machine vision enabled robotics technology. ReliVision is improving manufacturing efficiency by making work-object identification and inspections easier, smarter and unprecedentedly more accurate.
Application Engineer - Manufacturing Technology
Lockheed Martin Aeronautics
"Relimetrics is working with Lockheed Martin Aeronautics to develop AI-based machine vision enabled robotics for automating aircraft manufacturing and assembly processes. This innovation in automatic work-object identification allows the system to manage greater degrees of complexity than what is capable with traditional machine vision. Like Lockheed Martin’s other investments in advanced production technologies, the goal is speeding production and improving aircraft quality.”
Our customer is a global Tier I supplier of car seats. Meeting OEM standards for aesthetic quality requires a thorough quality inspection before seats are shipped to be installed in vehicles. Improving quality audit standards and creating full traceability of quality are key goals of its digitization strategy.
Today, quality inspections of car seats are done manually by operators, whose inspection is prone to error due to fatigue. Wrinkles and other car seat defects that are undetected or detected late lead to higher production costs due to rework and recalls.
There is high cost associated with each, and with increasing demand for quality, it is imperative for our customer to create full traceability of quality in its assembly to reduce costs and improve quality.
Relimetrics offers an industrial grade, shop floor ready system consisting of both hardware and software to digitize the quality of car seats.
Data acquired by the Relimetrics system are processed at the edge, providing evaluation of quality issues, such as wrinkles or other defects.
for a whole seat
Probability of detection
# of defect
At e.GO Mobile, Relimetrics implemented a very intuitive, flexible and efficient solution to capture product quality issues with Relimetrics‘ AI-based machine vision software, ReliVision, in partnership with HPE and PTC.
Quality inspections are becoming increasingly more complex in the automotive industry, given today’s accelerated production schedules and the proliferation of options and extras.
For an automotive OEM like e.GO, it is critical to adopt a quality digitization solution that is able to adapt to high production variability on the shop floor with high detection accuracy, and that does not rely on proprietary hardware enabling any industrial grade camera whether smart or conventional to be leveraged.
At e.GO, Relimetrics implemented a very intuitive and efficient solution to capture end-of-line quality issues using Relimetrics‘ AI-based machine vision software (ReliVision) in partnership with HPE and PTC.
ReliVision leverages existing industrial grade camera hardware, providing a flexible and scalable image inspection solution for anyone without any coding expertise to perform AI based machine vision and quickly deploy trained models in line for real time inspection on the shop floor and scale across inspection sites.
in scrap & rework
in scrap & rework
Our customer is a global Tier I automotive supplier of modular and assembled automotive parts. To keep OEM customers satisfied and prevent warranty claims, our customer is focusing on digitizing and improving product quality assurance.
Customer has already invested in embedded computer vision systems to digitize quality audit of drive shaft modules in its production line.
The return on investment has not been realized due to the deployed systems failing to deliver the required probability of detection (POD) and generating high percentage of false alarms.
As the module directly impacts vehicle safety, false detection is a significant liability.
Relimetrics augmented the existing camera hardware previously deployed at the customer, boosting the POD with its machine learning approach at the edge.
This solution enhanced and unlocked much more value from the previous investment.
It also fully integrated with the factory Manufacturing Execution System (MES), allowing our customer to use a single interface while using Relimetrics.
Increased existing PoD from
97% to 99.9%
of quality issues
Our customer provides automotive OEMs with high performance components with plastic parts, such as manifolds, ducts and cylinder head covers. As an early adopter of Industry 4.0 technology, it set out on a robust digitization journey in an effort to increase manufacturing efficiency and decrease costs.
Customer uses blow molding process to manufacture automotive filter components. Produced parts have to meet stringent customer quality demands and standards to avoid costly recalls.
Today, quality inspections are done manually by operators. Defective parts are either recycled or discarded. Recycling can lead to contamination, and discarding is costly and wasteful.
Process control is the main challenge. Currently, blow molding technicians adjust machine parameters to optimize production efficiency within quality specifications. Yet, quality drifts may go unnoticed for days.
With increasing quality demands, customer sought to make automated quality audit and process control an integral part of the production process.
Relimetrics deployed a fully-integrated system that digitized visual inspections of manufactured components. The system:
■ Inspects the dimensions and surface of every manufactured part and assures quality is within specifications
■ Correlates digitized quality data with machine and process data to optimize blow molding process
■ Identifies quality drifts in real-time and provides recommendations on how to adjust process parameters for optimized production, increasing productivity and throughput of a production line
Cost decrease related
to quality audit
Probability of detection
Our customer is a leading global supplier of formwork and scaffolding products in the construction industry. It is digitizing sorting and handling of returned articles in its rental businesses to have full traceability of quality before making decisions on article repair, reuse, and customer billing.
Our customer relies on manual sorting and inspection of articles returned to its rental sites and lacks ways to objectively assess damages. Damages caused by the renters are often overlooked, and renters are not charged for the incurred damage. Increasing number of articles in our customer's portfolio increases labor time and complexity, leading to mistakes in sorting and counting. This is amplified further with differences in contract terms and conditions with the renters.
Relimetrics provides a flexible and scalable image inspection solution, which can be managed directly by customers, for object recognition, counting and damage assessment. Interface modules enable connection to MES, ERP systems, enabling a closed loop inventory management process to be delivered to any production or rental facility. Relimetrics solution is hardware agnostic and can leverage existing hardware resources, allowing our customers to extract more value from their investment.
More accurate way to
determine root causes
Probability of detection
in inspection time
Covestro is a leading global supplier of innovative polymers used in nearly every part of modern life across industries and products. It is helping its own customers to innovate and create better products by designing better performing materials and modernizing manufacturing.
Strict quality demands in the construction industry puts high pressure on panel suppliers to deliver quality assured products.
Visual inspections are slow, subjective and often prone to errors, making defects undetectable until panels show signs of tear and wear while in use. Extraction of defective panels is costly, and expose the panel suppliers to fines and other liabilities.
Identifying the root cause of defects poses an additional challenge due to lack of understanding of correlation between machine, process and quality parameters.
Relimetrics solution digitizes quality inspection of polymer panels using ReliVision and provides a full overview of quality before the products leave the shop floor.
The system can be used to inspect a variety of quality parameters in the manufactured panels, including panel alignment, thickness, and surface anomalies, such as cavities and voids.
The deployed system also integrates with the shop floor MES, and offers the customer additional value metrics to improve its manufacturing and quickly detect quality drifts.
For more details, visit our Insight.tech article.
for a whole panel
Probability of detection
More accurate way to
determine root cause of defects
Relimetrics' Non-Destructive Testing (NDT) Module, RELI-NDT, digitizes visual inspections of defects in phased array ultrasonic testing (PAUT) as well as x-ray and infrared thermography inspections. Customers can define their specific requirements and train algorithms with well defined deep learning recipes without writing a single line of code.
There is a plethora of possibilities for things to go wrong in logistics operations and supply chains across industries. Take, for example, the wind energy industry. According to National Renewable Energy Laboratory, blade failure is one of the most common failure events in wind turbines, resulting in costly repairs and lost revenue.
The blades of a wind turbine must be able to withstand high winds in the field. As a result, the inspection of each blade, before getting installed on the field is a delicate process that requires the utmost precision. This inspection produces an enormous amount of data that needs to be evaluated to detect defects.
Today, the inspection of PAUT data is carried out by visual inspection, which is a time consuming and labor-intensive process prone to error due to fatigue.
Relimetrics’ AI accelerated NDT module, RELI-NDT, is designed to overcome the drawbacks of human visual inspection in PAUT, x-ray and infrared thermography inspections.
With RELI-NDT, customers can rapidly implement well defined AI algorithms to digitize visual inspection of PAUT data to inspect blade defects and their size. RELI-NDT can be configured to meet the specific requirements of ultrasonic inspection of any customer. A good example is renewable energy leader Siemens Gamesa, which engaged with Relimetrics to inspect its wind turbine blades with RELI-NDT.
Depending on the inspection requirements, RELI-NDT can perform inspection of adhesive joints on a single wind-turbine blade in 45 min. This is a major reduction in inspection time considering that the manual inspection consumes an average of 7 hours to complete a full blade inspection and requires 100% of the attention of the inspector during the shift.
Designed for high production variability manufacturing environments with a wide variety of configurations, RELI-NDT takes inspections of NDT data to a new level, providing a greater degree of accuracy than humans performing the tedious work.
RELI-NDT outperforms conventional inspection products relying on rule-based algorithms with its proprietary AI stack. Inspection algorithm scan be re-trained quickly to adapt to new circumstances and tolerance levels.
RELI-NDT provides well defined deep learning recipes, enabling anyone to train deep learning models without any AI or coding expertise.
RELI-NDT is also rapidly scalable within a manufacturer’s ecosystem: trained models and configurations can be managed with a centralized data management interface and shared across plants in the cloud.
Relimetrics’ Non-Destructive Testing (NDT) Module, RELI-NDT, digitizes visual inspections of defects in X-ray inspections as well as infrared thermography and phased array ultrasonic testing (PAUT). Customers can define their specific requirements and train algorithms with well defined deep learning recipes without writing a single line of code.
Radiographic testing is one of the most commonly used NDT methods to ensure reliability and safety of industrial products. However, today, the inspection of RT images is still mostly carried out by visual inspections as X-ray images often have poor contrast and are noisy, making it difficult to visualize anomalies.
Conventional radiographic testing based on human visual inspection is time consuming, labor intensive and prone to errors due to reading fatigue. This makes the reliable detection of defects one of the most challenging tasks in NDT. To overcome the drawbacks of human visual inspection, our customer seeks to digitize the inspection of X-ray images by leveraging advances in computer vision and AI.
With ReliVision, customers can rapidly implement well defined AI algorithms to digitize visual inspections of defects in X-ray data. ReliVision can be configured to meet the specific requirements of X-ray inspection of any customer. A good example is Asia’s biggest energy company Eneos, which engaged with Relimetrics to inspect anomalies in X-ray images with ReliVision.
Relimetrics provides a very intuitive and efficient solution to detect shades on X-ray images with a high detection probability under 1 second per image.
Using ReliVision, the defect inspection of X-ray data is automated, assessing the internal structure of manufactured components quality before they leave the shop floor.
Relimetrics solution serves as a standardized quality inspection automation platform that can be used across plants and ingest images from a variety of sources.
Probability of Detection
Inspection Time per Image
Positive Predictive Value