Latest / Internet of Things with Fexingo: Connected Devices, Sensors, and Industrial IoT / How IoT Sensors Are Tracking Carbon in Supply Chains
Transcript
- Lucas: So you've heard companies say their supply chains are carbon-neutral. But how do they actually know? That's where IoT sensors are stepping in, and the numbers are starting to get real. Luna: You mean physical sensors on shipping containers and factory floors tracking emissions directly, not just estimates? Lucas: Exactly. And that distinction matters because a lot of carbon claims today rely on industry averages — this product category uses X kilowatt-hours, therefore Y emissions. But a sensor on an actual assembly line capturing real-time energy draw tells a much more precise story. Lucas: Before we go deeper, I'll mention quickly — if you find conversations like these useful, listeners like you help keep this show ad-free and independent. You can find us at buy me a coffee dot com slash fexingo. No pressure, just an option to support what we do. Luna: Yeah, it's a small way to keep the podcast running without interrupting the flow with ads. Anyway — back to those sensors. What's a concrete example? Lucas: Let's talk about Bosch Rexroth. They're a German industrial supplier, part of the Bosch group. In 2025 they ran a pilot at their Homburg plant where they placed over 10,000 IoT sensors across their tier-2 supply chain — not just their own factory, but upstream suppliers making hydraulic components. Luna: 10,000 sensors — that's a lot of data points. What were they measuring? Lucas: Energy consumption per machine cycle, temperature in transit, vehicle fuel use for inbound logistics. The sensors fed into a dashboard that calculated carbon per component in near real time. What they found was that about 40% of the emissions their suppliers had been reporting couldn't be verified by actual meter data. Luna: That's huge — if those were the claims being used in their own ESG reports, that's a greenwashing risk. Lucas: Right. And Bosch Rexroth's goal was to cut that unverified gap. By the end of the pilot, they'd reduced the unverified portion to under 5% — not by changing operations, just by measuring better. The sensors themselves don't cut emissions, but they make the accounting honest. Luna: So it's a verification layer. How do you ensure the sensor data itself isn't tampered with? If a supplier wants to show lower numbers... Lucas: That's the second piece. Bosch Rexroth used a blockchain-anchored data pipeline — every sensor reading was hashed and recorded on a private ledger. So once a reading enters the system, you can't retroactively change it without breaking the chain. Luna: That adds credibility. But it also adds complexity — are there standards for how these sensors are calibrated across different environments? Lucas: Not yet, and that's a real hurdle. A temperature sensor on a truck in Arizona in July versus one on a factory floor in Norway — same model, different readings for the same true temperature. The industry is working on interoperability standards, but it's fragmented. The ISO is developing a framework for IoT-based carbon measurement, but it's still in draft. Luna: So early adopters like Bosch Rexroth have to build their own calibration protocols? That's expensive. Lucas: It is. But the alternative — using generic emission factors — is becoming less acceptable. Regulators are starting to push. The EU's Carbon Border Adjustment Mechanism requires importers to report embedded emissions. And California's climate disclosure laws, which take full effect in 2027, will require companies to report Scope 3 emissions — that's everything in the supply chain. Luna: And without IoT sensors, how do companies currently estimate Scope 3? Spreadsheets? Lucas: Largely. They ask suppliers to fill out surveys with average data. Or they use industry benchmarks from databases like Ecoinvent. Those are fine for rough orders of magnitude, but they don't capture real operational variance. A factory running on hydro power versus one on coal — same product, same process, wildly different emissions. Luna: So sensors give you actual granularity. What about the cost per sensor? Can a small supplier afford to deploy these? Lucas: Costs have come down dramatically. A simple energy-monitoring sensor that clips onto a circuit breaker costs about 30 to 50 dollars. A multi-sensor unit that tracks temperature, humidity, and vibration runs maybe 150. For a mid-sized factory, outfitting a production line might cost a few thousand. Over a year, that's negligible relative to the compliance risk. Luna: And what about data overload? 10,000 sensors generating continuous readings — how do you process that? Lucas: That's where edge computing comes in. Bosch Rexroth processed about 80% of the data locally on gateways, only sending summaries and anomalies to the cloud. That cut bandwidth costs and made the system responsive enough to flag a spike in energy use within minutes. Luna: So the edge does the heavy lifting, cloud stores the verified records. Makes sense. Are there other industries adopting this approach? Lucas: Logistics is the obvious next wave. DHL has been testing IoT pallet tags that record shock, temperature, and location. They're working with a Dutch startup called Sensolus that makes long-range, low-power tags. The idea is that by 2027 every pallet in DHL's European network could carry a sensor that reports carbon data along with tracking. Lucas: One specific pilot: DHL ran a trial on 500 pallets moving auto parts from Stuttgart to Valencia. They found that the actual transport emissions were 18% lower than the standard diesel-truck emission factor suggested, because the trucks used a mix of biodiesel and efficient routing. Without sensors, they'd have overreported. Luna: That's a good problem to have — reporting too high. But the reverse could happen too, right? A supplier using dirty fuel but claiming green. Lucas: Exactly. And that's where the sensor-based verification becomes a shield against accusations of greenwashing. If you can point to tamper-proof IoT data, your claim is much harder to challenge. Luna: What about the energy consumed by the sensors themselves? Does that offset any benefit? Lucas: It's negligible. A typical industrial IoT sensor draws milliwatts. Over a year, a thousand sensors might use less electricity than a single desktop computer. The carbon saved by accurate measurement and subsequent efficiency gains dwarfs the sensors' own footprint. Luna: Okay, so the tech is viable and the regulatory push is coming. What's the bottleneck to broader adoption? Lucas: Two things. First, data interoperability — getting sensors from different manufacturers to speak the same language. Right now, a Siemens sensor and a Honeywell sensor might output data in different formats. There's an industry group called the Open Carbon Data Alliance working on a standard schema, but it's early. Lucas: Second, the cost of integrating sensor data into existing ERP systems. Many companies run on legacy software that wasn't built to ingest real-time IoT streams. Retrofitting can be expensive, especially for smaller firms. Luna: Are there startups focused on that integration layer? Lucas: Yes. A company called CarbonChips offers a plug and play IoT gateway that connects to common sensors and pushes standardized carbon data into SAP and Oracle. They've raised about 20 million dollars and are piloting with a few European automakers. Luna: So the ecosystem is forming. Where do you see this in, say, three years? Lucas: I think by 2029, IoT-based carbon accounting will be standard practice for any company with serious Scope 3 reporting obligations. The cost of not doing it — regulatory fines, reputational risk, investor pressure — will exceed the cost of deploying sensors. It'll be like financial auditing: you can't just claim your numbers, you have to substantiate them. Luna: And sensors are the substantiation. Good place to leave it. Thanks, Lucas. Lucas: Thanks, Luna. For everyone listening, we'll be back next week with another angle on the connected world. Until then.