The next stage of industrial development requires machines that not only exchange data but also make autonomous payments. Here, a clear market trend meets an unresolved infrastructure problem: Industries are moving away from the rigid ownership of their means of production and relying on flexible, liquidity-preserving usage models. At the same time, highly automated micropayments via traditional payment methods like SEPA are hardly economically viable. Therefore, attention is turning to smart contract-enabled blockchain infrastructures.

Stablecoins and CBDCs (digital euros) are making headlines, but only one model is regulatory-compatible for European industry: Commercial Bank Money Tokens (CBMT) – tokenized bank money on a programmable infrastructure. Instant and precise enough to perform calculations to eight decimal places. And the sandbox project of the German Banking Association with Siemens and Evonik demonstrates that it works technically. Now the challenge is to scale up the concept and integrate it into industrial backbones.

Let's clear up a common misconception right away: speed isn't the issue. SEPA is already instant, finalized in seconds, and can be triggered fully automatically via EBICS or APIs. A single invoice based on a delivery note doesn't require CBMT. What traditional rail systems can't handle is micro-transaction volume: ten thousand payments of 0.3 cents each per day, each with a fixed overhead, simply don't make for a viable model. CBMT, however, moves value on a shared ledger. There, a microtransaction costs practically nothing. The eight decimal places are therefore not a gimmick, but a prerequisite for precisely pricing even the smallest units of service.

The viability of the model has been proven since May 2026. Siemens and Evonik conducted the first live transactions using CBMT, in a multi-bank setup with DZ Bank and Commerzbank. The pilot project initially demonstrates the regulatory and technical soundness of multi-bank processing. The high-frequency pay-per-use scenarios, for which the granularity is actually designed, are the logical next step. This brings within reach what has plagued Industry 4.0 so far: While it has networked machines down to the micro-detail, payment remains crude. This gap not only drives accounting to distraction, but also consumes working capital and hinders usage-based business models.

What Industry 4.1 means

If Industry 4.0 was networked production, then Industry 4.1 is the autonomous value chain, including payment. Machines not only communicate, they conclude contracts and settle them. This is driven by four developments: a machine-to-machine economy, in which the equipment itself becomes a market participant; pay-per-use models, which transform investments into ongoing operating costs; the tokenization of physical assets; and programmable supply chains.

The timing is no coincidence. Only now are a viable regulatory framework and technological maturity converging. Furthermore, usage-based procurement reflects a business trend: companies that bill for a machine per operating hour need a payment system that can bill per operating hour.

Three settlement options, only one carries

Three models are currently being discussed for the industrial pay grade. Media attention is inversely proportional to industrial suitability.

Stablecoins are liquid and globally usable, but they tie up additional capital and, above all, solve another problem: cross-border payments. Stablecoins from European providers are either new or still in their infancy. The current dominance of US-based providers poses a strategic risk, especially for European supply chains.

The digital euro is primarily a political project. Its functionality and programmability for industrial applications are unclear, and wider availability is not expected before 2029. For Industry 4.1 use cases currently being piloted, it will come too late.

Commercial Bank Money Token (CBMT) , or tokenized bank money from commercial banks, is the only option that is currently both regulatory and available. CBMT is balance sheet-relevant, deposit-protected, and legally embedded. In December 2025, BaFin formally classified CBMT as a deposit, not as e-money. This is an important detail, because its viability arises not from a new regulation, but from the application of a proven one.

From functional concept to functional layer

This model is also gaining international traction: In spring 2027, the largest US banks will launch a shared network for tokenized deposits via their own clearing houses. What's remarkable here, however, is not so much the "if" but the "how." Until now, each major bank operated its own blockchain: JPMorgan with JPM Coin, Citi with Citi Token Services, and BNY with its own service. The new network deliberately consolidates these individual initiatives onto a shared, cross-bank infrastructure, supported by more than a dozen institutions. This very step—away from the isolated, bank-specific structure and toward a shared layer—is the real news.

For industry, this is the key requirement. A factory pays suppliers who are held at a wide variety of banks. A CBMT that only functions within a single institution is worthless for an autonomous supply chain. Cross-bank interoperability is therefore not a convenience feature, but a fundamental business principle. Significantly, the German CBMT sandbox switched from a shared bank ledger to a bridge approach via the Universal Digital Payments Network for the same reason. Both sides of the Atlantic arrive at the same conclusion: Tokenized bank money only unfolds its value across bank boundaries.

The sandbox practice: Siemens, Evonik and the banking association

In Germany, this infrastructure is already under development. Testing is being conducted for M2M payments, conditional payments, pay-per-use for industrial plants, automated supply chains, and offline M2M payments. Participating banks include Commerzbank, Deutsche Bank, DZ Bank, and UniCredit; while on the industrial side, Siemens, Evonik , BASF , Airplus, and Mercedes-Benz are involved, along with Festo in an offline pilot project. Technology partners include Giesecke+Devrient, GFT Technologies, UDPN , and Oracle.

The development proceeded in stages: from the Proof of Concept in 2024 with ten participants, through the sandbox launch in November 2025 in Frankfurt and the pre-production trials from February 2026, during which the CBMT bridge switched to the UDPN approach, to the first live transactions by Siemens and Evonik in May 2026. This completes the transition from sandbox to productive pilot.

The path to widespread market readiness remains open. Three layers are vying for control over the bottleneck: the technical settlement latency, integration into the ERP landscape, and the identity layer. DZ Bank recently estimated the timeframe for market-ready adoption at six months to two years.

The final piece of the puzzle: Digital identity

When a machine orders and pays autonomously, a question arises that the financial layer alone cannot answer: Who is actually acting here, and with what authority? The good news is that industry doesn't need to invent its own procedure for this. Europe already has the standard, and it's called eIDAS 2.0. For individuals, there's the EUDI Wallet; for organizational identity, there's the EU Business Wallet, including powers of representation and authorizations. This will not only simplify digital supply chains—keyword: Digital Product Passport ( DPP )—but also ensure legally secure automated digital payments.

For companies, this is less a question of architecture than of risk. Without this layer, autonomous payment doesn't offer efficiency gains, but rather a loss of control with liability implications: Who authorized the machine, up to what amount, and in whose name? The eIDAS 2.0 stack transforms this vulnerability into a regulated, auditable process. And because banks are already building this identity layer for their own EUDI obligations by 2027, the real investment isn't the initial setup, but rather the bridge from the bank's eIDAS identity layer to the company's IT infrastructure, from verified authorization to the triggered payment. This interface between regulated financial identity and operational technology is the necessary architectural work now.

CMBT intergration: from machine event to settlement

What companies can do now

The focus is now shifting from the strategic to the operational level. The Tech Radar Verdict is at the "Assess" stage. The model is ready for pilot testing, but not yet for production. This classification is based on the BaFin classification as a deposit from December 2025 and the first live transactions from May 2026.

Initial pilot projects within closed value chains featuring high transaction frequency and pay-per-use-enabled systems can quickly demonstrate integration capabilities and benefits. A business relationship with one of the following banks is also required: Commerzbank, Deutsche Bank, DZ Bank, and UniCredit.

The leverage is the tied-up capital in the supply chain: Pay-per-use transforms investments into ongoing costs, continuous billing shortens payment terms, and regulated bank money is daily business as opposed to stablecoins.