Physical AI, High-NA Chipmaking and Digital Infrastructure Resilience — September 8, 2026
Automation History does not claim authorship or ownership of the third-party articles featured in our Daily Articles section. All articles, research, and reporting are credited to their original authors and publishers, with links provided to the original sources. Automation History shares these links solely for educational and informational purposes, helping readers discover and learn from relevant content. All rights remain with the respective authors and publisher
1. MIPS introduces three workload-specific physical-AI platforms
MIPS announced September 7 three open, modular platforms intended to move AI into machines, vehicles and embedded systems. Acies targets edge inference, Actus supports real-time motion control, sensor fusion and predictive maintenance, and Aegis addresses low-latency processing in safety-critical systems.
Why it matters: Separating inference, real-time control and safety workloads can make systems easier to validate and troubleshoot. Repair teams will still need clear fault boundaries, version records and safe fallback behavior when AI hardware interacts with conventional controls.
Read the September 7 Robotics 24/7 report
2. ASML adapts High-NA lithography for larger AI chips
ASML said September 8 that it is working with major chipmakers to accommodate larger data-center processors on its newest High-NA EUV tools, including through larger photomasks. The change could improve system productivity by approximately 40%, although production adoption is expected to occur gradually between 2028 and the early 2030s.
Why it matters: Larger masks and finer features will increase demands on contamination control, calibration, precision motion and supplier support. Maintenance capability must mature alongside the process rather than being added after high-volume production begins.
Read the September 8 Reuters report
3. South Korea forecasts a major power increase for AI and chip production
South Korea expects AI data centers and semiconductor factories to add 25–30 gigawatts of electricity demand, Reuters reported September 8. The government is evaluating nuclear expansion as Samsung and SK Hynix accelerate a semiconductor cluster backed by roughly $590 billion in planned investment.
Why it matters: Automated production depends on generating capacity, substations, cooling and grid reliability as much as production equipment. Leaders should incorporate utility constraints and power-quality risks into every large automation expansion.
Read the September 8 Reuters report
4. Firmus secures OpenAI as an anchor customer for Malaysian capacity
Nvidia-backed Firmus signed a multiyear agreement to supply OpenAI from two planned Malaysian data centers, Reuters reported September 8. The agreement takes Firmus’ contracted capacity beyond 900 megawatts as Southeast Asia weighs AI investment against electricity and water constraints.
Why it matters: Data centers are industrial facilities with pumps, cooling loops, switchgear, controls and backup systems. Rapid expansion needs maintainable designs, documented redundancy and technicians capable of restoring service without compromising customer workloads.
Read the September 8 Reuters report
5. Corporate boards prepare for digital-infrastructure disruption
A Capgemini survey reported September 8 found that 86% of 1,300 executives saw significant exposure to digital-sovereignty risks, while 65% had budgeted countermeasures. Nearly half said replacing a critical technology provider would take three to twelve months, and more than one-third expected it to take longer.
Why it matters: Industrial continuity planning should include controller software, cloud services, AI models and remote-support platforms—not only physical spare parts. Vendor substitution must be tested before a geopolitical or cyber incident removes access.
Read the September 8 Reuters report
6. Mistral raises €3 billion to expand European AI
Mistral announced September 8 that it raised €3 billion at a valuation of approximately €21 billion, or $24 billion—the largest private European technology equity round to date. The company plans to finance frontier-model research as it approaches a projected $1 billion in annual recurring revenue.
Why it matters: A stronger European AI supplier may improve deployment and data-residency options, but rapid vendor growth does not eliminate dependency risk. Industrial users should retain model portability, exportable data and a tested alternative for essential workflows.
Read the September 8 Reuters report
7. New sensor tackles difficult hydrogen-flow measurement
Researchers described September 7 a meter that divides flow into four channels and can be inserted into high-pressure pipes of different sizes and wall thicknesses. Simulations indicate it can measure hydrogen mass accurately even after a 90-degree upstream bend, although certification is still pending.
Why it matters: Reliable measurement is essential for process control, equipment efficiency, leak detection and commercial accounting. Its certification stage is a useful reminder that encouraging laboratory performance is not yet equivalent to dependable plant-floor instrumentation.
Read the September 7 NTNU research summary
8. Cyberattack disrupts Boston Scientific’s global manufacturing
Boston Scientific disclosed September 8 that a cybersecurity incident disrupted manufacturing and order shipments worldwide. The medical-device company now expects a material effect on its third-quarter and full-year 2026 results and does not anticipate meeting its previous sales and adjusted-profit forecasts.
Why it matters: This connects cyber resilience directly to production output and financial performance. Plants need isolated backups, manual shipping contingencies, restoration priorities and drills that test whether manufacturing can recover without spreading a compromise.
Read the September 8 Reuters report
9. AI will make 70 years of broadcast history searchable
Reuters and ITN announced September 8 that more than one million historical video clips will become available through Reuters Connect beginning January 1, 2027. AI will add transcripts, shot descriptions and translations, while another 12,000 hours of footage will be digitized.
Why it matters: For Automation History, this demonstrates how AI can convert an inaccessible physical archive into searchable operational knowledge. The same approach can make old maintenance videos, commissioning records and retired-equipment documentation useful again.
Read the September 8 Reuters–ITN announcement
10. Ancient mold provides direct evidence of repeatable Jerusalem production
Archaeologists reported September 7 that an eighth- to sixth-century B.C. clay mold from Jerusalem’s Qishla complex is the city’s first direct evidence for local production of Judean figurines. Craftspeople formed heads with molds and made bodies separately, creating a repeatable assembly process.
Why it matters: For Automation History, the mold records an early form of standardized production involving reusable tooling, separate components and controlled replication. It also illustrates why manufacturing tools can reveal more about a process than surviving finished products alone.