IAP · Pulse 6 of 8 · Agentic Demand Family v3 · monthly composite
Industrial / Physical Agent Pulse
The third wave of agentic AI — agents that move atoms, not just bits. Factory automation, robotics, machine vision, industrial-CAD with virtual-twin AI, embedded silicon for sensor fusion, edge gateways for remote deployment, automotive autonomous-vehicle stack. Six sub-baskets, monthly composite (industrial-CAPEX-cycle is 5-15 years; faster cadence would be noise). LSCC explicit here for low-power-FPGA on the embedded edge — moved out of CEP per v3 spec because FPGA is industrial/control, not consumer.
21 points · since 2026-05-27
Multi-Basket Equity Gaps
Six sub-baskets organized by industrial-vertical. Heavy international representation (Siemens DE, ABB CH, Schneider FR, Fanuc/Yaskawa/Keyence JP, Advantech TW). Embedded Semis is the leading-indicator sub-basket — when physical/industrial agents start showing up, FPGA + embedded NPU + sensor-fusion silicon picks up first.
| Sub-Basket | Vertical | Weight | 21d Basket Return | Score | Tickers |
|---|---|---|---|---|---|
| Embedded Semis | Low-power FPGA + embedded MCU/NPU for sensor fusion + industrial control | 20% | +1.13% | 55.6 | LSCC · NXPI · STM · ADI · TXN |
| Industrial Automation | Factory-automation order intake + agentic-overlay platforms | 20% | -1.08% | 44.6 | ROK · EMR · SIE.DE · ABBN.SW · SU.PA |
| Robotics / Machine Vision | Robotic-controller shipments + machine-vision components | 20% | -3.12% | 34.4 | CGNX · 6954.T · 6506.T · TER · 6861.T |
| Industrial Software | Industrial-CAD + Simulation + Virtual-Twin with agentic-AI overlays | 15% | +24.25% | 100.0 | PTC · SNPS · DSY.PA |
| Edge Gateways | Rugged edge-server adoption for factory + remote AI deployment | 10% | -8.50% | 7.5 | DELL · HPE · 2395.TW |
| Auto / Physical AI | Automotive AI design-wins + autonomous-vehicle stack | 15% | -6.25% | 18.7 | QCOM · NXPI · STM · MBLY |
Earnings-NLP module breakdown
| Module | Weight | Raw signal | Score | Status |
|---|---|---|---|---|
| Industrial automation orders | — | signal 65/100 | 64.9 | live · 2026Q3, 2026Q2, 2026Q1 |
Evidence — ROK 72/100 · EMR 62/100 · SIE.DE 74/100 · SU.PA 52/100 ROK 2026Q3 scaling · accelerating confidence: high
“Q3 sales came in above our expectations. Reported sales were up 8% and organic sales were up 10%...E-commerce & Warehouse Automation sales were up 30% year-over-year...Semiconductor delivered another strong quarter...Data center remained a ” EMR 2026Q3 scaling · accelerating confidence: high
“Underlying orders grew 7% in the third quarter, with broad-based growth across all business groups, led by Software & Systems, which was up 10%... Our project funnel grew $1.2 billion to $12.4 billion, up 8% year-over-year.” SIE.DE 2026Q2 scaling · accelerating confidence: high
“DI's automation revenue was up by 6% to EUR 3 billion, led by the short-cycle factory automation business... April shows another strong growth, even above our forecast — automation globally. China industrial market is recovering.” SU.PA 2026Q1 early · steady confidence: medium
“Industrial Automation delivered growth above 4%...broad-based recovery in Discrete Automation. ARR in software was up 12%.” Extracted from company earnings-call transcripts (AlphaVantage for US; Yahoo/Quartr for EU/JP) via Claude Sonnet, temperature 0, confidence-damped. Diary entry on our process, not investment advice. | ||||
| Robotics + machine-vision | — | signal 58/100 | 56.6 | live · 2026Q2, 2026Q1 |
Evidence — CGNX 52/100 · TER 42/100 · 6954.T 68/100 · 6506.T 68/100 CGNX 2026Q2 early · steady confidence: medium
“For this application, Cognex vision systems will be mounted on robots to inspect fully assembled server racks and confirm that all major components are installed correctly and meet strict quality requirements.” TER 2026Q2 early · steady confidence: medium
“Robotics revenue was $100 million, up 33% year-over-year and 9% quarter-over-quarter. Electronics manufacturing and semiconductor revenue increased by 50% from Q1 and is now the largest end market segment in this group, which includes AI da” 6954.T 2026Q1 scaling · accelerating confidence: medium
“The 'projects involving several thousand units' mentioned in Q3 have steadily been converting into orders...orders have exceeded our initial expectations, and we believe a portion is due to the favorable assessment of Physical AI.” 6506.T 2026Q1 scaling · accelerating confidence: high
“Shipments of robots for semiconductor applications are expected to reach an all-time high this fiscal year and remain extremely strong... More than 200 units of MOTOMAN NEXT are reportedly under deployment.” Extracted from company earnings-call transcripts (AlphaVantage for US; Yahoo/Quartr for EU/JP) via Claude Sonnet, temperature 0, confidence-damped. Diary entry on our process, not investment advice. | ||||
| Auto / physical-AI design-wins | — | — | 50 (neutral) | stubbed |
The three Earnings-NLP modules together carry 25% of the composite (sub-baskets 75%). Industrial-automation-orders and robotics-shipment-velocity now read live from earnings-call transcripts (Rockwell, Emerson, ABB, Siemens, Schneider; Fanuc, Yaskawa, Cognex, Teradyne); auto-design-wins remains stubbed. Expand “Evidence” for the underlying metrics and verbatim quotes.
Methodology
IAP is Pulse 6 of 8 — the industrial/physical agent layer. The Closelooknet v3 spec calls this the "third wave": structurally enormous TAM (trillions), but slow adoption-curve and fragmented public data. IAP is more watchlist than active-trading-signal — monthly cadence reflects that industrial-CAPEX cycles are 5-15 years, not quarters.
Six sub-baskets weighted Embedded Semis 20% + Industrial Automation 20% + Robotics/Vision 20% + Industrial Software 15% + Edge Gateways 10% + Auto/Physical AI 15%. Stock-price layer pulls from Closelooknet Data Lake. The three Earnings-NLP modules (25% of composite) are broken out in the module table above — industrial-automation-orders and robotics-shipment-velocity read live from earnings-call transcripts.
Cross-pulse-cross-read: IAP↑ + LSCC/NXPI/STM↑ = physical/industrial agents show up in embedded semis FIRST. IAP↑ + EIP-flat = edge-driven industrial AI, less cloud-dependent — LSCC/STM bull setup. Note: Synopsys (SNPS) now includes Ansys post-2024-acquisition; the industrial-software basket reflects that consolidation.
Pulse reading is a diary entry on our process, not investment advice.
FAQ · from the current data · as of 2026-10-19
Quick answers
What does the Industrial / Physical Agent Pulse measure?
IAP tracks the "third wave" of agentic AI — agents that move atoms, not just bits — across factory automation, robotics, machine vision, industrial CAD virtual-twin AI, embedded sensor-fusion silicon, edge gateways and the automotive autonomous-vehicle stack, on a monthly composite.
What is the current IAP composite reading?
As of Oct 19, 2026, 08:01 AM UTC, the IAP composite reads 49.3 out of 100, a Scaling momentum regime, across 25 covered tickers.
How are the six IAP sub-baskets weighted?
Embedded Semis, Industrial Automation and Robotics/Vision each carry 20%, Industrial Software 15%, Auto/Physical AI 15%, and Edge Gateways 10%.
Why is the IAP composite monthly rather than daily or weekly?
Industrial capex cycles run 5 to 15 years, so faster cadences would be noise. IAP is built as more of a watchlist than an active-trading signal, tracking a structurally large but slow-moving addressable market.