The Early Scale: AI Data Centers Can't Wait for the Grid as Manufacturers Lack the Plumbing

Good morning

Good morning. It's Wednesday, July 15, and the theme of the day is infrastructure under pressure. Data centers are eating construction budgets, AI pilots are stalling on the factory floor, and logistics costs just told us volatility is no longer a weather event, it's the climate. If you're running a business that touches physical infrastructure, digital systems, or supply chain, today's edition is aimed directly at you. Grab the coffee. Let's go.

The Big Three

AI Is Starving the Grid, and Fast Storage Is the Stopgap

AI data centers demand power in months; grid upgrades take years. The gap between those two timelines is now a critical infrastructure risk. CBAK Energy's 26650 V2.0 LFP cells just entered customer validation for AI data center backup power applications after a 15-month R&D program, and flexible fast storage is emerging as the near-term fix while utilities scramble to catch up. FERC is stepping in, but the structural mismatch between AI's power appetite and grid construction timelines isn't going away soon.

The B2B angle: If your facilities roadmap includes a data center build or expansion in the next 18 months, budget for on-site battery storage and backup power as baseline infrastructure, not optional insurance.

93% of Manufacturers Have MES. Only 23% Have It Talking to the Rest of the Business.

Rockwell Automation surveyed 1,560 manufacturing decision-makers and found that while MES (Manufacturing Execution System) deployment is nearly universal, enterprise-wide integration remains rare, just 23% have fully connected it. A separate Kaufman Rossin survey found 73% of manufacturers are still stuck in AI pilot mode, and the core blocker is the same in both cases: legacy ERP systems and siloed data. The tools exist. The plumbing doesn't.

The B2B angle: Before buying another AI or automation layer, audit whether your MES and ERP can actually share data, the Rockwell numbers suggest there's a 70-point gap between having the system and using it effectively.

Logistics Volatility Is Structural Now, the 2026 State of Logistics Report Says Stop Waiting for It to Pass

U.S. business logistics costs fell to $2.4 trillion in 2025, down from prior highs, but the 2026 State of Logistics Report identifies five structural forces, not a demand cycle, reshaping the operating environment. Meanwhile, a split is widening between consumer and industrial sectors: consumer sentiment is softening while parts of the industrial economy hold firm, forcing supply chain leaders to plan two separate strategies simultaneously. The era of one-size planning is over.

The B2B angle: Supply chain leaders should build separate demand-planning models for consumer-facing and industrial-facing segments right now, the divergence isn't temporary, and a single forecast is a liability.

Also worth knowing

B2B ecommerce is no longer a pilot program. MSC Industrial's Q3 sales crossed $1 billion with ecommerce leading the channel mix, Kawasaki Engines USA grew average order value 500% through digital investment, and Global Industrial posted 9.2% Q1 growth. The companies that treated digital channels seriously years ago are now collecting the compounding returns.

Siemens Energy is rebranding as Omterra, consolidating Siemens Gamesa and its grid operations under one name as it moves toward full independence from Siemens AG. For enterprise energy procurement teams, this is a supplier identity change worth tracking, contracts, contacts, and compliance documents tied to the old brand will need updating.

Enterprise operators are moving toward industry-specific cloud platforms as generic solutions increasingly fall short on vertical compliance and workflow needs. Industry clouds are gaining real traction, not just as a vendor pitch, but as a procurement pattern shift. If your tech stack evaluation still starts with horizontal platforms, it may be time to rerun that search.

Smart plays for the week

Audit your MES-to-ERP data flow this week and map exactly where the handoffs break before your next automation or AI investment. The Rockwell Automation data shows a 70-percentage-point gap between MES deployment and full integration, adding AI on top of disconnected systems compounds the problem, not the payoff.

If you're marketing to manufacturers or industrial buyers, lead your next campaign with integration outcomes, not feature lists, 'we connect your existing stack' beats 'we have AI' right now. With 73% of mid-market manufacturers stuck in pilot mode due to data infrastructure gaps, buyers are primed to respond to solutions that fix the plumbing, not add more tools.

Add on-site battery storage and backup power as a line item in any data center build or expansion RFP you're issuing in the next 90 days. The grid-to-AI-demand timeline mismatch is structural, CBAK Energy's LFP cells entering customer validation and FERC intervention signal that utilities cannot be the only plan.

Something to think about

Every planning assumption that treats current disruption as an aberration rather than a baseline is now a strategic liability. The companies winning in logistics are the ones that stopped waiting for normal to return.

Teach me something: Manufacturing Execution System (MES)

An MES is the software layer that sits between your high-level business systems (like ERP) and the actual machines on your factory floor. It tracks and controls work-in-progress in real time, scheduling jobs, monitoring equipment, recording quality data, and managing labor on the line. Think of it as the operating system for production. The problem most manufacturers face today isn't that they lack an MES, Rockwell's data shows 93% have one, it's that the MES isn't talking to the rest of the business. When your ERP and MES don't share data, you end up making procurement, scheduling, and AI decisions based on information that's already stale by the time it arrives.