Latest / Internet of Things with Fexingo: Connected Devices, Sensors, and Industrial IoT / How IoT Sensors Are Preventing Dockworker Injuries at Ports
Transcript
- Lucas: So earlier this year, the Port of Rotterdam reported something that made me stop and take notice: in 2025, they deployed a network of vibration sensors and wearable tags across one of their largest container terminals, and within twelve months, dockworker injuries dropped by 34 percent. Luna: Thirty-four percent — that is a big number. Was this a pilot program or a full-scale rollout? Lucas: It started as a pilot on Terminal B, which handles about two point five million TEUs a year — that's twenty-foot equivalent units, the standard container measure. They equipped around 400 workers with wearable tags that broadcast their location in real time, and they embedded vibration sensors on the dock itself, near the crane rails and the automated guided vehicles. Luna: So the vibration sensors — what exactly are they picking up? Are they detecting ground movement from approaching equipment? Lucas: Exactly. Those sensors measure low-frequency vibration patterns — the kind that come from a straddle carrier or a reach stacker moving within a certain radius. The system cross-references that data with the wearable tag locations. When a worker is detected within, say, three meters of an active heavy vehicle, both the worker's tag and the vehicle operator receive an audible and haptic alert. Luna: Haptic — so the wristband buzzes. That makes sense in a noisy environment where you might not hear an alarm. But I imagine the real challenge is false positives. A busy dock is basically constant motion. Lucas: That was the biggest hurdle during the first three months. The system was flagging near-misses dozens of times per shift — most of them were harmless, like a worker walking past a stationary crane. So the team had to tune the algorithm to ignore equipment that's powered down or idling for more than thirty seconds. They also introduced a geofence around 'safe zones' — marked areas where workers can cross without triggering alerts. Luna: Did they have to adjust the physical layout of the terminal too, or was it purely a software fix? Lucas: A bit of both. They repainted the safe zones and added physical barriers to guide foot traffic in the highest-density areas. But the real breakthrough came when they started using the sensor data to redesign the workflow. For example, they noticed that most of the false alarms clustered around a particular corner where containers were stacked — it turned out that the vibration from a forklift turning that corner was being misinterpreted as a head-on approach. Once they relocated two sensor nodes and adjusted the angle, the false positive rate dropped by 80 percent. Luna: That's a great example of how the technology is only as good as the deployment. So the 34 percent reduction — was that mostly prevention of serious injuries, or did it include minor incidents like slips and trips? Lucas: The 34 percent refers specifically to lost-time injuries — meaning incidents that caused a worker to miss at least one full shift. Total reported incidents, including minor ones, dropped by 41 percent. But the most dramatic reduction was in what they call 'struck-by' incidents — a worker being hit by a moving vehicle or load. Those fell by 57 percent. Luna: Fifty-seven percent. That is genuinely life-saving. Before we go deeper, I want to take a quick beat here because this kind of conversation — concrete numbers, real-world impact — it's exactly why we keep this show ad-free. We don't run commercials, we don't have sponsors, and we want to keep it that way. If this kind of tech deep-dive is valuable to you, the simplest way to support that choice is at buy me a coffee dot com slash fexingo. It's a small gesture that lets us stay independent and focused on the substance. Lucas: Yeah, absolutely. And it makes a real difference — it's what lets us spend time on stories like this one. So back to Rotterdam: the cost side. The entire pilot — 400 wearables, 120 vibration nodes, plus the software and integration — came in at around 1.2 million euros. The port estimates that each lost-time injury costs them roughly 45,000 euros in direct expenses — medical, compensation, overtime for replacements. Based on the injury reduction, they calculate a payback period of about fourteen months. Luna: Fourteen months is fast. And that's just the direct costs — it doesn't include avoided litigation, regulatory fines, or the hit to worker morale. So the business case is solid. But I wonder — is this technology scalable to smaller ports? Rotterdam is one of the biggest in the world, with a lot of resources. Lucas: That's a fair question. The good news is that the unit costs are dropping fast. Two years ago, a wearable tag with haptic feedback and a battery life of six months cost about 250 euros. Today, it's closer to 90. Vibration sensors have come down similarly. The Port of Ghent in Belgium is running a scaled-down version with just 60 wearables and 30 sensors, targeting their container yard. Their initial data shows a 22 percent reduction in near-misses after only six months. Luna: So the pattern seems to be: start with a high-risk area, prove the ROI, then expand. What about the privacy angle? Workers wearing location trackers — that can feel intrusive. Lucas: Rotterdam took a very deliberate approach. They involved the labor union from day one. The tags only transmit location data — no audio, no video, no biometrics. And the data is anonymized for analysis; supervisors can't see real-time locations of individual workers unless an alert is triggered. They also gave workers the ability to turn off the tag during breaks — it disables the location broadcast but keeps the haptic receiver active, so they still get alerts. Luna: That seems like a reasonable compromise. I also read that they're planning to integrate with the terminal's automated guided vehicle system — so the AGVs themselves can slow down or reroute when a worker is nearby. Is that part of the next phase? Lucas: It is. That's phase two, scheduled for later this year. The idea is that instead of just alerting the human operator, the AGV will automatically reduce speed from 20 kilometers per hour to 5 kilometers per hour when a tagged worker enters a 10-meter zone. The port's safety director told me they expect that to reduce struck-by incidents by another 30 to 40 percent on top of the current gains. Luna: That's the kind of sensor fusion that really unlocks safety. So what's the biggest lesson from this case that applies to other industrial environments — not just ports, but warehouses, factories, construction sites? Lucas: I think the biggest lesson is that the technology is no longer the bottleneck. The sensors are cheap enough, the battery life is long enough, the connectivity is reliable enough. The real challenge is organizational: getting buy-in from workers, tuning the false positive rate, and integrating the data into existing safety protocols. Rotterdam spent almost as much on change management and training as they did on hardware. Luna: So it's a sociotechnical system, not just a tech installation. That's a good note to end on. Thanks, Lucas. Lucas: Thanks, Luna. And for anyone curious about the specifics, the Port of Rotterdam published a white paper on the pilot in April — it's openly available on their website. We'll link it in the show notes.