If you work in an industrial business today, you’ve probably felt the shift. Sustainability reporting used to mean pulling together a broad picture once a year, a corporate-level snapshot for a report nobody outside the sustainability team read too closely. That’s no longer enough. Frameworks like the Corporate Sustainability Reporting Directive (CSRD), the Methane Act, Life Cycle Assessment (LCA) standards, and various ISO certifications now ask for something much more specific: real-time, product-level accountability. Regulators, customers, and auditors increasingly want to know the environmental footprint of a specific batch, a specific machine, a specific product, not just the company as a whole.
Here’s the part that surprises most people: the bottleneck usually isn’t a lack of data. Industrial sites already generate enormous amounts of it - sensor readings, production logs, maintenance records, energy meters. The real problem is context. That data sits in disconnected systems that were never designed to talk to each other, so nobody can easily answer a question like “what was the carbon footprint of this specific batch of product?” without weeks of manual digging.
This is one of the problems Cognite Data Fusion is built to solve. It pulls information out of the systems where it’s normally trapped in the control systems that run the plant floor, the business systems that track orders and inventory, even engineering drawings such as piping and instrumentation diagrams and maintenance records, and connects it all into a single, unified map of how the operation actually works. Think of it less as a data warehouse and more as a live, connected model: a sensor reading is linked to the machine it came from, which is linked to the production run it was part of, which is linked to the product that eventually shipped. Once that connection exists, turning raw activity into a defensible, audit-ready number stops being a weeks-long research project every time someone asks for one.
Emissions at the product level, without the spreadsheets
Increasingly, it's not just regulators asking for this level of detail - it is our customers. A growing number of buyers now expect, or are mandated by suppliers to report the carbon footprint of the specific product they're purchasing, not a company-wide average, because that number feeds directly into the buyer's own supply-chain emissions reporting. Historically, meeting that request meant estimating with historical averages or manually stitching together spreadsheets, slow, easy to get wrong, and hard to defend if a customer or auditor pushes back. When energy consumption and material inputs are linked to the specific production run that made a given batch, that number can be generated directly from what actually happened on the factory floor, instead of an average applied after the fact. That distinction has real commercial weight: a defensible, verified carbon figure - the kind that can sit behind an Environmental Product Declaration - is simply a stronger claim to make in a market where more buyers are starting to ask the question at all.
Passing the audit, faster
CSRD’s “limited assurance” requirement and standards like ISO 14067 (which governs how you calculate a product’s carbon footprint) both boil down to the same ask: can you show your working? An auditor needs to trace a headline number - say, “2.3kg CO2e per unit” - all the way back to the raw sensor data it was built from, not just take your word for it. When that trail already exists as a natural byproduct of how the data is connected, verification goes from a weeks-long forensic exercise to something an auditor can check in a fraction of the time. That’s a real, practical benefit: less time spent proving your numbers are real, and more confidence that a sustainability claim on a product - like an Environmental Product Declaration - will hold up to scrutiny rather than becoming a liability.
Environmental performance as a quality metric, not an afterthought
ISO 14001 (environmental management) and ISO 9001 (quality management) are usually treated as separate compliance boxes to tick. But the same continuous-improvement logic (plan, do, check, act) runs through both, and it works far better when it’s fed by real-time data instead of a once-a-year audit. Industrial energy management is a good illustration of what this looks like in practice: software that pulls in electricity, gas, water, and steam data across an entire site down to individual machines, and links it directly to production output, so a plant can see energy used per tonne or per unit made, not just a total on a utility bill. That per-unit number makes it possible to catch an inefficient process before it shows up as an unexplained spike in next quarter’s report, and to prove with real data that an efficiency gain didn’t come at the cost of an undisclosed environmental slip.
The bigger point
None of this is really about satisfying a regulator for its own sake. The companies that can produce this kind of granular, trustworthy data get something more valuable: the ability to make a sustainability claim about their own products and back it up with the same rigor they’d apply to any other critical business metric. That turns compliance from a once-a-year scramble into a standing capability - and increasingly, into a genuine point of difference with customers who are asking harder questions about where their products come from, and their impact.
Key
CSRD – Corporate Sustainability Reporting Directive
EUDR – EU Deforestation Regulation
LCA – Life Cycle Assessment
SCADA – Supervisory Control and Data Acquisition
ERP – Enterprise Resource Planning
P&IDs – Piping and Instrumentation Diagrams
EPD – Environmental Product Declaration
PDCA – Plan-Do-Check-Act
KPI – Key Performance Indicator