Probing of What AI has Learned Continues
Last week, I pointed out a future bottleneck to industry adoption of advanced AI systems is neural network explainability. Right on cure, AI leaders Google and Amazon released blog posts on techniques to help understand the patterns encoded in trained neural networks. Google asks, “Do Wide and Deep Networks Learn the Same Things”? While Amazon adapts a “technique for removing confounders from causal models, called instrumental-variable analysis, to the problem of concept-based explanation”. It’s still a leap to figure out the eventual impact of this work for heavy industry and manufacturing environments. But one thing manufacturers could consider is using ‘shallow’ neural networks. FICO explains the benefits of shallow AI such as more easily telling “regulators how their AI system functions because it’s easier to probe smaller neural networks.” For now, predictive maintenance and visual inspection applications will continue to dominate the AI systems deployed into the factory since humans can augment any unexplainable gaps in the AI for these use cases using conventional methods.
Visual Inspection
Complex Fabrications & Intricate Parts
Acoustic Monitoring
Also, give a listen to The Digital Thread’s Increasing Speed and Flexibility for Customers: The Protolabs and 3D Hubs Story episode.
Assembly Line
Machine learning optimizes real-time inspection of instant noodle packaging
During the production process there are various factors that can potentially lead to the seasoning sachets slipping between two noodle blocks and being cut open by the cutting machine or being packed separately in two packets side by side. Such defective products would result in consumer complaints and damage to the company’s reputation, for which reason delivery of such products to dealers should be reduced as far as possible. Since the machine type upgraded by Tianjin FengYu already produced with a very low error rate before, another aspect of quality control is critical: It must be ensured that only the defective and not the defect-free products are reliably sorted out.
F-16s Are Now Getting Washed By Robots
The Wilder Systems solution actually leverages technology previously developed for robotic drilling in commercial aircraft manufacturing and converts these components and subsystems into an automated washing system. The main changes have involved the development and addition of robot end-effectors to provide the water and soap spray, waterproofing of the robots themselves, and a robot motion path, which is dependent on the type of aircraft to be cleaned.
Amazon Lookout For Equipment – Predictive Maintenance Is Now Mature
Amazon Lookout for Equipment is designed for maintainers, not data scientists, and it comes from a place of knowledge. Incorporating expertise and insight gleaned from maintaining its own assets, Amazon aims to make it as easy as possible for users to get started and begin seeing value, addressing potential issues around usability and configurability.
In terms of technical abilities, it currently only covers simple assets like motors, conveyors, and servos – essentially, the kind of assets Amazon itself uses. It doesn’t yet monitor more sophisticated assets like robots or CNC machinery, although, in time, I do not doubt that these, too, will also be covered. As it stands, though, it will be competent for a lot of standard factory equipment.
John Deere and Audi Apply Intel’s AI Technology
Identifying defects in welds is a common quality control process in manufacturing. To make these inspections more accurate, John Deere is applying computer vision, coupled with Intel’s AI technology, to automatically spot common defects in the automated welding process used in its manufacturing facilities.
At Audi, automated welding applications range from spot welding to riveting. The widespread automation in Audi factories is part of the company’s goal of creating Industrie 4.0-level smart factories. A key aspect of this goal involves Audi’s recognition that creating customized hardware and software to handle individual use cases is not preferrable. Instead, the company focuses on developing scalable and flexible platforms that allow them to more broadly apply advanced digital capabilities such as data analytics, machine learning, and edge computing.
Exploring Additive Manufacturing Opportunities: Optimizing Production with Hyundai Lifeboats
This project was the epitome of Explore. Just as myself, Director of Innovation at Materialise, and others from the Mindware team, had no experience or knowledge of producing lifeboats, the Hyundai team was unaware of the capabilities and limitations of 3D printing. So, the first step in this project was bringing our two worlds together to pinpoint a relevant business challenge for Hyundai Lifeboats that we believed could best be solved via additive manufacturing.
Easier said than done. We dove into an interactive workshop session in which we discovered each side’s perspectives, expectations, and blind spots. We first discussed how AM could increase the boat’s value — with enhanced speed, performance, functionality — but we were met with hesitancy from the Hyundai team.
Google Cloud and Seagate: Transforming hard-disk drive maintenance with predictive ML
At Google Cloud, we know first-hand how critical it is to manage HDDs in operations and preemptively identify potential failures. We are responsible for running some of the largest data centers in the world—any misses in identifying these failures at the right time can potentially cause serious outages across our many products and services. In the past, when a disk was flagged for a problem, the main option was to repair the problem on site using software. But this procedure was expensive and time-consuming. It required draining the data from the drive, isolating the drive, running diagnostics, and then re-introducing it to traffic.
That’s why we teamed up with Seagate, our HDD original equipment manufacturer (OEM) partner for Google’s data centers, to find a way to predict frequent HDD problems. Together, we developed a machine learning (ML) system, built on top of Google Cloud, to forecast the probability of a recurring failing disk—a disk that fails or has experienced three or more problems in 30 days.
Surge Demand
Just not good enough, ‘just in time’ supply chain practices have been upended due to the pandemic and other freak events. Altair takes a look at when digital twins are coming through a poll of 124 manufacturing professionals. Large-scale Space manufacturing is coming sooner than you think. As the electric vehicle era is upon us, a former Tesla engineer is looking to battery recycling as a lucrative business.