No Fatigue in Industrial AI Agents

This week: AI agents sequence machine processes, create fatigue models, embody intelligent humanoids, and copilot CNC machines. Manufacturing history back to the Model A and tie-bar-less machines.

Aug 17, 2025

∙ Paid

Shop Talk

Capturing this week's zeitgeist

The 1928 Ford Model A's debut at the Rouge plant marked the automotive industry's first major pivot from pure manufacturing efficiency to consumer-centric personalization, establishing a pattern that continues today with AI integration. While Henry Ford's Model T offered utilitarian transportation "in any color so long as it's black," the Model A introduced meaningful choice through multiple body styles, hydraulic brakes, electric starters, and safety features—representing a fundamental shift toward designing vehicles around human needs rather than production constraints. This progression from mechanical personalization options in the 1920s to today's AI-driven systems that learn individual driving patterns and preferences illustrates how manufacturing technology has consistently evolved to meet increasingly sophisticated consumer demands. Just as the Rouge plant's integrated manufacturing enabled the Model A's variety while maintaining efficiency, modern AI systems now optimize production schedules, manage supply chains, and customize user experiences in real-time, proving that the most successful technology implementations, whether the Model A or today's BEVs, succeed by putting human needs first and then applying innovation to meet those evolving expectations. Welcome to Industry 5.0.

The Ford Model A was a groundbreaking vehicle that represented a turning point in the history of automobiles. The Model A’s success was a testament to the changing tastes and demands of consumers in the automotive market. — The Ford Model A's popularity lives on today!

Up next: Learn about the state of the robotics industry at RoboBusiness, which takes place October 15-16 in Santa Clara, California or K, the international trade fair for innovations in the plastics and rubber industry, October 8-15th in Düsseldorf, Germany.

Assembly Line

This week's Industry 5.0 breakthroughs and frontier technologies of the built world.

How Engel Broke the Mold at K in 1989

✍️ Author: Norbert Sparrow

Engel reports that more than 85,000 tie-bar-less machines have been installed in plants around the world so far, and that number continues to grow as it refines that innovation. In fact, at this year’s K show in October, Engel will introduce an electric model with new features. But let’s not get ahead of ourselves. What made tie-bar-free design a milestone in plastics processing to begin with?

Why did we build a usage-based fatigue model using GenAI technology?

There are several ways to build fatigue models. The traditional way is physics based, where methods such as multi-variate analysis and finite element analysis are used to model fatigue. This can be very complex and time consuming, and some physics are simply too difficult to model. Today, machine learning is increasingly used to build models. Machine learning is data driven, so as long as there are enough data to represent fatigue, the training process will figure out the model. However, machine learning has its own shortcomings, such as noise, bias, over/under-fitting (i.e. prediction accuracy), and care must be taken to build an accurate and robust machine learning model.

This was when we turned to Generative AI (GenAI) models. Remember we mentioned that the Random Forest model worked on the cumulative operational statistics? GenAI models worked directly on the real-time operational data. This allowed us to capture much more details in the data, and one benefit of this was that we needed fewer failures to train the model. To be clear, the total amount of data absorbed by the GenAI model was vastly increased, because now we are ingesting 1-second data, but the number of failures needed for model training was significantly reduced.

How Trihedral shares SCADA data

Supervisory control and data acquisition systems (SCADA) play a critical role in collecting process data from across distributed systems and making it available across an organization. But what if you need to share that data beyond your firewall? What are the keys to doing so efficiently and safely?

What Is Industrial DataOps & Why Does Every Manufacturer Need It? /LNS Research/
Manage Production With Asset-Tracking Technology /Assembly/ by implementing wireless asset-tracking technology from Sensolus.

🦾🤖🧠 Embodied Intelligence: The PRC’s Whole-of-Nation Push into Robotics

✍️ Author: Sunny Cheung

Beijing is mounting a coordinated campaign to get ahead in next-generation artificial intelligence (AI) hardware through a nationwide surge in robotics. While companies in the West like Tesla and Boston Dynamics introduced physical AI to global audiences years ago, the People’s Republic of China (PRC) is now rapidly assembling an impressive array of competitors, marshalling industrial, academic, and financial resources to scale up its new national champions. The race is well underway.

Read more at The Jamestown Foundation

China's Industrial Robot Installations Rose 5% Last Year Amid Global Decline /Yicai/
Fulin Precision Soars After Placing China’s First Large Order for Embodied Industrial Robots /Yicai/
Robots run, punch and score at World Humanoid Robot Games in China /The Guardian/
ShengShu Technology launches Vidar multi-view physical AI training model /The Robot Report/

🗜️🧠 DeepMS: A data-driven approach to machining process sequencing using transformers

✍️ Author: Jaime Maqueda, David W. Rosen, Shreyes N. Melkote

Efficient and intelligent machining process sequencing remains a key challenge in computer-aided process planning (CAPP). Traditional methods often rely on manually defined rules and explicit feature recognition, limiting their adaptability across diverse parts and evolving manufacturing environments. Recent advances in deep learning (DL), particularly in transformer-based sequence modeling, offer a promising alternative by enabling systems to learn sequencing logic directly from data without explicitly modeling complex rules. This paper presents a novel DL framework that predicts machining sequences directly from the 3D geometry of final parts. Operating on voxelized representations, the model generates an ordered sequence of machining operations, each associated with a volumetric shape representing the material removed from raw stock—eliminating the need for predefined features or rule-based logic. The framework integrates a transformer-based sequence autoencoder to model operation order and an encoder based on 3D convolutional neural networks (CNN) to map final part geometry to sequence representations. To efficiently handle high-dimensional voxelized data, a 3D CNN autoencoder is employed to compress voxelized removal volumes. Components of these pretrained models are combined into an inference pipeline that generates machining sequences directly from the final part geometry.

Read more at Journal of Manufacturing Systems and GitHub

🖨️🧠 Artificial Intelligence for Control in Laser-Based Additive Manufacturing: A Systematic Review /IEEE/
🧠 Phi-Reasoning: Once again redefining what is possible with small and efficient AI /Microsoft Research/

We started covering ‘copilots’ way back in January 2024, help your friends and colleagues get an early signal on emerging industrial technology by sharing the publication.

Share Exponential Industry

New Product Introduction

Highlighting new and innovative facilities, processes, products, and services

🗜️🧠 Mastercam Copilot: The Future of AI-Powered CNC Programming

Unlike traditional user support or static help documentation, Mastercam Copilot is a centralized, intelligent support system that boosts productivity, reduces onboarding time, and makes using Mastercam easier than ever.

🗜️🧠 Generative AI Catalyzes Growth in Industrial Robotics & CNC Machining /Machine Design/
🗜️🧠 What CNC Taught Us About Adopting AI in Manufacturing /Modern Machine Shop/
🧠 Market Insight: 10 GenAI-Powered Industrial Copilot Vendors To Watch In 2025 /Verdantix/

💿 U.S. Startups, Multibeam and SecureFoundry, Debut First E-Beam Tools for Wafer Writing

✍️ Author: Alan Patterson

Two U.S. startups, Multibeam and SecureFoundry, are making the world’s first e-beam lithography tools that write chip designs directly to a silicon wafer without using masks. ASML needn’t worry about losing its top spot in the litho business, however. The startup e-beam toolmakers say their gear will enhance rather than replace the EUV and DUV equipment that only ASML currently makes.

In exclusive interviews with EE Times. Multibeam president Ken MacWilliams, SecureFoundry CEO Lex Keen, and industry veteran Aki Fujimura spoke about the outlook for e-beam.

🚙🔋 CATL Launches New Repair Service for CTP Batteries to Cut EV Owners' Costs

Contemporary Amperex Technology, the world's biggest maker of electric vehicle batteries, has rolled out a repair service for its batteries using cell-to-pack technology to reduce expenses for car owners and seek a new growth driver.

CATL's aftermarket unit Ning Service launched the CTP repair service on Aug. 10, the Ningde-based parent company announced on the same day. The new service is significantly cheaper than full pack replacement, saving substantial costs for users and breaking the industry's "replace-only, no repair" deadlock for CTP batteries, it noted.

Business Transactions

This week's top funding events, acquisitions, and partnerships across industrial value chains.

🌬️ Orsted plunges as it seeks $9.4 billion to cope with Trump's hostility to wind power /Reuters/

🇦🇺 NRN secures $67m in one of Australia's largest climate tech Series A deals

NRN (National Renewable Network), Australia's leading distributed energy infrastructure platform, has finalised a $67.2 million Series A funding round, one of the largest ever in Australian climate tech. The raise includes equity and debt, with major support from a superannuation-linked fund.

NRN enables energy retailers like Alinta Energy (an NRN-partner retailer) to offer customers solar and battery systems with no upfront costs, repayments, or system maintenance fees. With NRN, retailers can quickly build a distributed renewable energy portfolio and white-label it as their own Virtual Power Plant plan.

🇨🇦 BinSentry Raises $50 Million to Modernize Animal Feed Supply Chains with AI-Powered Sensor Technology

BinSentry, a technology company helping major agriculture enterprises optimize feed supply chains through AI-powered sensors and real-time monitoring, announced a $50 million Series C funding round led by Lead Edge Capital. This investment will support the acceleration of BinSentry’s global expansion, following a year of extraordinary performance and growth.

BinSentry combines AI-powered sensors and proprietary software to bring real-time visibility and automation to the animal feed supply chain. Its solar-powered, self-cleaning sensors monitor on-farm feed levels with 99% accuracy, eliminating the need for manual inventory checks and preventing costly feed outages. Paired with a mobile dashboard, the technology allows feed mills and large agricultural businesses to better forecast demand, reduce waste, and improve profitability. BinSentry supports some of the leading animal agriculture companies in North America and Brazil, including Cargill, Wayne-Sanderson Farms, and Hanor.

🇺🇸 Squint Raises $40M Series B to Expand Agentic Manufacturing and Accelerate Human Potential with AI

Squint, the leading Manufacturing Intelligence Platform, announced it has raised $40 million in Series B funding, bringing total funding to more than $59 million. This round was led by The Westly Group and TCV, with repeat investors Sequoia Capital and Menlo Ventures.

With this latest investment, Squint will advance its AI capabilities to lead the Agentic Manufacturing revolution, deepen integration across manufacturing workflows, and expand to new sectors including energy, logistics, and field services. The funding also positions Squint to meet surging demand. For example, one Fortune 50 customer is currently expanding to over 70 additional sites, while a Global Fortune 500 manufacturer is rolling out Squint to 10,000 field techs.

🇺🇸 Softwear Automation Secures $20 Million in Series B1 Funding

Softwear Automation, the pioneer in autonomous sewing technology, announced the successful close of its $20 million Series B1 funding round, marking a major milestone in its mission to revolutionize sewn goods manufacturing. The round was led by a strategic investment from BESTSELLER, the Danish fashion company behind brands such as JACK & JONES, VERO MODA, and ONLY. The investment was made through Invest FWD, BESTSELLER's dedicated innovation and investment platform for advancing sustainability in the fashion industry. Current investors (including CTW Venture Partners, SRI Capital, MacDonald Ventures) also participated in the financing.

As part of the collaboration between Softwear Automation and BESTSELLER, CFO Thomas Børglum Jensen, on behalf of BESTSELLER, joins the board of Softwear Automation. This at a time of rapid progress towards commercialization of Softwear's fully automated SEWBOT® worklines.