Improving industrial performance increasingly depends on information held by suppliers, infrastructure operators and other businesses. Research from the World Economic Forum examines why that information remains difficult to share, even where the potential benefits are substantial.
Manufacturers have become increasingly capable of measuring what happens inside their own operations, yet some of the information that could make the greatest difference to performance remains beyond their control. It may sit with a supplier planning deliveries, an energy provider managing capacity or a neighbouring business with resources that could be put to better use. A factory can improve its internal forecasts considerably and still make poor decisions because it cannot see what is happening elsewhere.
This is the challenge examined in the World Economic Forum’s August 2026 report, Intelligent Industrial Ecosystems: Turning Proximity into Intelligence, developed with Accenture. Its central argument is that the next gains from industrial digitalisation may depend less on what individual companies can extract from their data than on whether independent organisations can use selected information together. That creates commercial and organisational difficulties which better analytics alone will not resolve.
The report cites a European Commission estimate that around 80 percent of industrial data in the EU goes unused. The figure does not establish how much of that information would be commercially valuable, but it gives an indication of the gap between collecting industrial data and making productive use of it. Within companies, that problem is familiar. Across business boundaries, it becomes more complicated because the organisation holding the information may have little reason to release it.
The economics of sharing
A supplier may be able to improve a manufacturer’s production planning by revealing more about its capacity and constraints. From the supplier’s perspective, however, that same visibility could weaken its position in price negotiations or expose commercially sensitive details to a much larger customer. The combined operation might become more efficient without the supplier receiving a fair proportion of the benefit. Unless that imbalance is addressed, the rational commercial decision may be to share as little as possible.
The Forum identifies competitive sensitivity, unequal bargaining power and unclear incentives as persistent obstacles to industrial data collaboration. Companies must agree not only what information can be exchanged, but who funds the arrangements, what each participant is entitled to do with the data and how benefits will be divided. These questions become harder when organisations have different regulatory obligations or when the same companies cooperate operationally while competing in other markets.
Differences in digital maturity create further obstacles to collaboration, with the report showing that enterprise resource planning adoption among EU businesses in 2023 ranged from 37.9 percent for small enterprises to 86.3 percent for large ones. Although ERP adoption is not a direct measure of data-sharing capability, the disparity illustrates why imposing one technical model across an entire supply chain may prove unrealistic. Smaller organisations cannot necessarily commit the engineering resources, system upgrades or specialist staff that a large manufacturer takes for granted.
For that reason, proposals for an industrial data ecosystem cannot begin with an assumption that everyone will contribute freely to a common platform. They need a credible answer to a more basic question: what does each participant stand to gain, and what control will it retain? Where a project offers a collective saving but imposes costs or risk unevenly, technology may make collaboration possible without making it commercially acceptable.
What shared visibility can change
The Port of Rotterdam offers a practical example of why this matters. Its Starlings project, launched in 2025, allows industrial businesses in the port cluster to coordinate the exchange of utilities including electricity, steam and hydrogen. A company able to reduce demand can potentially agree compensation with another that benefits from changing supply, creating a more flexible arrangement than fixed exchanges between individual sites.
The port authority acted as an independent coordinator and committed €2 million to a three-phase pilot. Importantly, the platform initially shared flexibility bids rather than the underlying sensitive operational information. The early trials delivered efficiency gains of around 5 percent, according to the report. Subsequent modelling suggests more sophisticated data exchange and system-wide optimisation could more than triple the original gains, with larger benefits possible if participants also invest in shared infrastructure. Those further returns are projections, not results already achieved.
There is a considerable difference between improving the management of existing assets and identifying investments that no individual company could justify alone. If several plants can demonstrate complementary patterns of demand, infrastructure providers have stronger evidence for developing shared capacity. Better data can therefore influence decisions about physical assets, but the value only materialises when participants can agree how investment and operational responsibility will be shared.
A different illustration comes from Kalundborg in Denmark, where 17 public and private organisations exchange more than 35 streams of water, energy, materials and residual products. The industrial symbiosis has developed over decades, with surplus resources from one facility becoming inputs for another. Its shared database makes these flows visible across the partnership, helping identify opportunities that would remain obscure if each business considered only its own processes.
The Forum reports that the cluster conserves approximately four million cubic metres of groundwater annually and recycles more than 200,000 tonnes of residual materials. Those achievements cannot simply be credited to AI: the database currently relies on people to interpret the relationships and act on them, while automated matchmaking remains an area being explored. That qualification is important because the immediate benefit comes from organisations having a common view of potential exchanges, rather than from the sophistication of the algorithm examining them.
At the Ordos-Envision Net-Zero Industrial Park in China, meanwhile, shared information supports the coordination of renewable generation, energy storage and industrial demand across more than 50 enterprises. The report attributes an approximately 10 percent reduction in overall energy costs to this coordinated approach. Sensitive operational information remains within individual companies, with aggregated or less sensitive outputs contributed to the common platform. Here, the ability to preserve local control is part of what makes system-level optimisation workable.
Control matters more than centralisation
These cases point towards an approach in which companies share enough information to enable a decision without surrendering every underlying dataset. The Forum distinguishes centralised arrangements, which pool information for shared analysis, from federated arrangements, where data stays with its owners and approved outputs are exchanged. Hybrid models combine elements of both and, the report argues, have become the most common approach among leading industrial ecosystems.
A manufacturing example is Factory+, developed by the University of Sheffield’s Advanced Manufacturing Research Centre. Its architecture can give participating facilities separate instances and local control while allowing selected machine telemetry or analytical outputs to be shared through agreed interfaces. Shared data definitions help different systems interpret those contributions consistently. At the AMRC’s COMPASS composites facility, which has Boeing as its first user, selected values can be shared with the wider research environment without giving that environment unrestricted access to all facility data.
That level of selectivity is essential, but it does not remove the need for governance. Organisations still require enforceable agreements covering permitted use, responsibility for data quality, access rights and what happens when a participant leaves. There must also be a way to resolve disputes when the savings realised by one company depend on investment or information supplied by another. A neutral coordinator, such as a port authority or industry body, can help establish the rules, although it cannot eliminate conflicts of commercial interest.
The security implications are equally substantial. Linking industrial businesses can create new pathways through which a compromised supplier, shared service or data connection affects other participants. The Forum argues that collective resilience requires agreed security baselines and coordinated incident response, rather than relying entirely on the defences of individual firms. Every proposed connection needs to be assessed for the operational benefit it brings and the additional exposure it creates.
The report consequently recommends starting with a defined value case, building the necessary governance and technical arrangements, then measuring the outcomes before expanding. That is more demanding than launching a broadly conceived data platform, but it provides a practical test of whether participants have a reason to cooperate. A narrowly specified exchange that improves scheduling or reduces energy costs may be worth more than a large collection of data with no agreed purpose or commercial owner.
Industrial companies have spent years trying to make better use of the information they already possess. The Forum’s research suggests the next limitation may be information they cannot access, or information their partners cannot afford to share on the proposed terms. Overcoming it will require manufacturers to treat data collaboration as a commercial agreement with technical requirements, rather than a technical project that will somehow produce agreement. The most valuable industrial intelligence may depend on the willingness of several companies to make decisions together, while retaining good reasons to trust the arrangement.
