A U.S. robot factory says it beat China's PCB lead time

1. A robot-run U.S. factory says it builds circuit boards in 4 days, at China prices

Four days, not nine. That's the average build-and-ship time CircuitHub says its automated circuit board factory in South Deerfield, Massachusetts, now hits, compared with a typical quick-turn shop in China. The company announced on October 26 that the factory, called the Grid, has reached price parity with similar Chinese shops and is expanding capacity fivefold, with a UK site planned for 2027. Levitating movers carry each board between robots that print solder paste, place parts, solder and inspect, and AI routes every board on its own path, so there are no changeovers between designs. CircuitHub says one unit each of 100 designs runs as easily as 100 units of one design, and that it has shipped more than 2.2 million boards. Why it matters: CircuitHub says 95% of electronics projects involve fewer than 10,000 units, and those small, high-mix orders are where offshore lead time and freight hurt most. Price parity is the company's own claim, so test it with a quote.

So what for buyersFor low-volume, high-mix board assemblies, put one automated domestic quote next to your current offshore source and compare landed cost and total lead time, not unit price alone.

You are a procurement analyst choosing which circuit board assemblies to re-quote with a faster domestic source. Inputs, all anonymized, one row per assembly: part code, annual quantity, typical order quantity, orders per year, current unit price, current lead time in days (order to dock), expedite and premium freight cost in the last 12 months, late deliveries in the last 12 months.
[PASTE TABLE]
Assumption: cost per late delivery [AMOUNT].
1) Flag assemblies with a typical order quantity under [10,000] units.
2) For each flagged row, estimate the yearly cost of waiting: expedite and premium freight plus late deliveries times the cost per late delivery. Show your math for the first row.
3) Rank the flagged assemblies by cost of waiting, then by lead time.
4) Recommend the top 3 to re-quote and the one fact to confirm before moving each.
Output: a table with columns part code, order quantity, lead time, cost of waiting, rank; then the top 3 with one sentence each. Keep it under 300 words. Use anonymized data only. List any missing or inconsistent data instead of guessing.

Sources: CircuitHub announcement (via EMSNow) Company announcement · ASSEMBLY Independent reporting

2. Oracle launches AI agents that draft allocation plans and hunt maverick spend

Twenty-five new agentic applications became available across Oracle's Fusion cloud apps on October 26, and three are aimed at supply chain teams. Oracle says Supply Planning Allocation compares scenarios for splitting scarce materials and capacity across competing demand, Shipment Load Planning looks for load consolidation to cut transport cost, and Procurement Spend flags maverick spend and contract leakage. According to Oracle, each one prepares plans or recommendations that people review and approve.

So what for plannersIf you plan on Oracle Fusion, ask your admin which of these you can use, and write down your allocation rules before an agent starts proposing its own.

Sources: Oracle announcement Company announcement · Diginomica Independent reporting

3. TSMC weighs up to 20% higher prices for AI chip packaging

Six to eight months. That's how long equipment lead times for advanced chip packaging have stretched, up from three to six, as TSMC weighs raising prices for its CoWoS packaging by 10% to 20%, Taiwan's Commercial Times reported, citing supply chain sources. CoWoS joins AI processors to high-bandwidth memory in data center accelerators, and demand has spread from GPUs to server CPUs and custom chips. TSMC has not confirmed the plan. TrendForce, citing the Liberty Times, says TSMC has also set wafer price increases of 3% to 6% starting January 2027.

So what for procurement leadersIf you buy AI servers or accelerator hardware for data centers, assume packaging costs keep rising into 2027 and ask suppliers now how much of an increase they plan to pass through.

Sources: TrendForce Independent reporting · Yahoo Finance Independent reporting · Investing.com Independent reporting

4. GXO takes Digit humanoid robots to two more U.S. warehouses

More than 100,000 totes. That's how many Agility Robotics' Digit humanoid has moved at GXO's Flowery Branch, Georgia, warehouse under a robots-as-a-service deal signed in 2024, GXO said on October 26. GXO now plans to put Digit to work on repetitive tote handling at e-commerce sites in Statesville, North Carolina, and Lebanon, Tennessee, starting in early 2027. The companies say Digit has logged more than 65,000 hours of real-world operation across all of its deployments.

So what for warehouse managersHumanoid pilots are turning into multi-site contracts. If you're evaluating one, ask the vendor for live-site numbers like totes moved and hours run, and price it per task.

Sources: GXO announcement Company announcement · Logistics Manager Independent reporting