Adoption of AI-based Machine Vision Is Growing


Machine vision technology and applications continue to hold a spot in my heart even though it became less interesting to me over the years. My $25K system from 1993 went to $10K by 1996 to $5K and under by 2000, all the while becoming more powerful and easier to use. However, I’ve lately talked with a company working on Artificial Intelligence for machine vision. It seems to be getting some traction. Here are two recent releases.

Landing AI and the Association for Advancing Automation (A3) have released a report from a survey covering a range of topics, including the level of adoption, benefits, and challenges of implementing AI-based visual inspection. 

“As evidenced by the survey, AI-based machine vision is already creating value for the manufacturing industry, with proven benefits including improved accuracy, flexibility, and reduced cost,” said Dongyan Wang, Vice President of AI Transformation at Landing AI. “The availability of easy-to-use AI tools specifically designed for visual inspection will drive further industry adoption and bring the benefits of AI to more organizations.” 

The report from Landing AI and A3 found that companies have a high degree of confidence in the effectiveness of AI, with 55% of respondents saying their overall opinion is either high or very high. The survey showed that 26% of respondents have adopted AI-based machine vision while 41% say they plan to in the future. Of those who are using AI, improved accuracy is the top benefit (62%).

While implementing AI, the primary challenge is scarcity of data on which to train AI models, noted 62% of manufacturers. Companies were also concerned about scalability as solutions moved out of pilot programs, with 27% saying they struggled when moving from proof-of-concept to initial deployment. 

Additional findings from the Landing AI and A3 report include:

  • In a heavily automated sector, manual inspection is still playing an important role, with 40% saying their inspection is either completely or mostly manual.
  • The confidence level of businesses regarding AI effectiveness is high with 26% saying they are already using AI for visual inspection.
  • When it comes to using AI, scarcity of data, complexity of integrating AI within existing infrastructure, and the inability to achieve lab results in production are the top three challenges. 
  • Most businesses prefer to have ownership of their AI projects either by developing in-house or by working with a vendor. 

AI Visual Inspection Platform to Improve Quality and Reduce Costs

Meanwhile, Landing AI unveiled LandingLens, an end-to-end visual inspection platform specifically designed to help manufacturers build, deploy, and scale AI-powered visual inspection solutions. 

“AI-powered visual inspection solutions have demonstrated clear advantages over conventional methods, yet the overall adoption is slow as many companies get stuck after some small-scale proof-of-concept projects,” said Andrew Ng, founder and CEO of Landing AI. “LandingLens, developed with the know-how and expertise from building and shipping many visual inspection projects, is best positioned to bridge the gap and enable teams to succeed. With an intuitive interface, Landing AI’s visual inspection platform allows even non-AI experts to train and deploy a model with a few clicks, putting the ownership of the solution in the hands of users so they can build and update the solutions without being beholden to a third-party AI team.”

Features

  • Defect and Data Management: A robust data preparation module helps customers improve efficiency and produce more accurately labeled data, resulting in better performing models in less time. 
  • Model Iteration: Training and evaluation tools enable users to rapidly develop a deployment-ready model. 
  • Continuous Learning: Inference and monitoring modules allow users to scale their AI solutions by enabling them to deploy production-ready models to the edge with ease, while continuously monitoring those deployments from a central location.

Benefits

  • Reduced Project Lifecycle: Reduces the AI project development life cycle time by up to 67%.
  • Cost reduction: Reduce the overall AI development and deployment costs by up to 60% 
  • Improved Accuracy: Improves accuracy of Machine Learning models through more accurately labeled data
  • Scaling: Enable users to effectively scale to multiple production lines across many factories

TRUMPF and SICK Develop First Industrial Quantum Sensor

In a continuation of Sensor Day at The Manufacturing Connection, here is an announcement about an Industrial quantum sensor development.

  • Cooperation agreement on development of industrial quantum sensor signed
  • Successful functional test of the world’s first quantum optical sensor for serial production
  • First sensors in industrial use from 2021
  • Experts estimate total market for quantum sensors at about EUR 1.1 bn. by 2023

The wholly owned TRUMPF subsidiary Q.ANT and sensor specialist SICK will in future work together on the development of quantum optical sensors. Representatives of the two high-tech companies signed a cooperation agreement in early November 2020 to make quantum technology for sensors available for industrial use. 

Quantum sensors enable measurements with unprecedented accuracy. The signing was preceded by a successful functional test of the world’s first quantum optical sensor for serial production. 

“SICK is expanding its position as a worldwide technology leader in the sensor sector by embarking on the production of quantum sensors. Quantum sensors are a key technology for the future of industry,” says Robert Bauer, Chairman of the Executive Board of SICK AG. 

As a market and technology leader for industrial sensors, SICK will be responsible for application development and sales of the product. As a specialist in quantum technology, the TRUMPF subsidiary Q.ANT takes on production of the measurement technology – and thus the sensor’s core technology. 

“Quantum technology is an enormous opportunity for German and European industry. This will be the first time that the partnership between our two high-tech companies will involve a product for serial production. The quantum sensor enables highly accurate measurements and will provide insights that will lead to completely new industrial applications,” says Peter Leibinger, Chief Technology Officer at TRUMPF. The first use of the new quantum sensors is planned for 2021.

Highly precise measurements of the smallest of particles 

Quantum sensors have hitherto mainly been used in research. For the first time, Q.ANT and SICK have now successfully completed functional tests for an industrial application. 

“Quantum technology enables, for example, ultrafast measurements of the movement and size distribution of particles. With industrialization of these sensors, not just us but Germany – as a high-tech location – takes a major step towards the commercialization of quantum technology,” says Michael Förtsch, CEO of Q.ANT. 

Using laser light, quantum sensors permit highly efficient measurements that would be impossible with conventional processes. 

“Quantum technology is the next level for sensors because it shifts hitherto firmly established technical limits. Using quantum effects, additional details can be perceived from signal noise where, up to now, no specific signals would have been measureable. This enables the measurement of particles that are about two hundred times smaller than the width of a human hair,” says Niels Syassen, Senior Vice President R&D at SICK and responsible for the project there. The quantum sensors will initially be used for analyzing substances in air.

Industrialization leads to increased market volume

Quantum sensors could in future become everyday equipment in various industries: For example, they could be used in civil engineering to visualize underground structures before construction work begins; in the pharmaceutical industry they could make it easier to determine the best composition of tablet powder; in the electronics sector circuits could be inspected through surfaces; and highly accurate measurements could be made in industry in general. The market for quantum sensors could grow steadily with industrialization. Experts at Germany’s National Academy of Science and Engineering (Acatech) estimate the worldwide market volume of industrial quantum sensors at about EUR 1.1 bn. by 2023.

New Series of Cloud-Based, Wireless Sensors

Swift Sensors Launches Series 3 – The First IoT Platform With One-Press Activation, Predictive Alerts and 6-8 Year Battery Life

This is going to be “sensor day” at The Manufacturing Connection. I have two interesting new products. Sensors, of course, form the foundation for the Industrial Internet of Things.

Swift Sensors calls itself a cloud-based wireless IoT company. It has launched its new Series 3 wireless sensor system with improvements in wireless range, battery life, durability and ease of deployment.

Series 3, re-engineered from the ground up, incorporates cutting-edge BLE5 technology with the Nordic nRF52840 SoC’s on-chip cryptography, multiprotocol communication, and power management.

Key features:

  • Three times the sensor range, up to 300 feet, for use in large facilities.
  • Four times the number of sensors (up to 150) that each gateway can securely support, for broader and more detailed sensor coverage. 
  • Extended sensor battery life of six to eight years, five times the previous version, to reduce maintenance and sensor downtime.

“We designed our industrial sensor system from the ground-up to be simple to deploy and use,” said Ray Almgren, CEO of Swift Sensors. “Our new sensors and gateways will drive mainstream adoption of IoT in the industrial markets. Customer feedback over the past several years has also driven configuration, deployment, and durability improvements in the wireless system.”

For example, the first batch of 12 Series 3 sensors includes a predictive vibration sensor that adds peak-to-peak, RMS, and crest factor monitoring. This lowers the entry cost of industrial equipment predictive maintenance (PdM) for customers who want to reduce their downtime and maintenance expenses. 

Other improvements in the Series 3 platform include:

  • A water-resistant (IP 67) sensor enclosure for extra durability in harsh environments. 
  • “Instant On” one-touch sensor activation for immediate measurements with zero configuration needed.
  • “Find My Sensor” console button that blinks a green LED on the deployed sensor to identify it.
  • An internet status LED that shows if the gateway is connected to the internet, connected locally or not connected.
  • Enhanced web app and LED indicators that simplify gateway Wi-Fi setup.
  • An optional USB cellular network module for deployments when ethernet and WiFi are not available. 

“Using advanced BLE5 communication protocols we have developed a world-class data delivery system to optimize range, reliability, and security of the Series 3 sensors,” said Dean Drako, founder and executive chairman of Swift Sensors. “Our cloud-based wireless monitoring system now offers a breadth of common measurements using the latest IoT technology.”

Swift Sensors plans to release additional Series 3 sensors soon to measure sound, light, pressure, motion and the presence of harmful gasses. Swift Sensors Series 3 components are available on the company’s website and through the Swift Sensors global reseller network.