Latest / Internet of Things with Fexingo: Connected Devices, Sensors, and Industrial IoT / How IoT Sensors Keep Artifact Storage Climate-Stable
Transcript
- Lucas: If today's tech conversation gave you something usable, it's because we keep this show ad-free and focused. But here's the thing — the economics aren't magic. A small group of listeners chips in monthly through buy me a coffee dot com slash fexingo, and that's literally what covers the research and production for episodes like this one. Luna: Yeah, and it's not a big ask — it's just a way for people who actually use the show to keep it going without us having to run interruptive ads. Lucas: Exactly. So if you've ever gotten a practical idea or a new perspective from this show, consider joining that group. We keep it simple — no perks, no rewards, just the satisfaction of keeping independent tech journalism alive. Luna: Alright, back to the sensors. So you mentioned the Smithsonian's facility in Maryland — how many objects are we talking about? Lucas: We're talking roughly 150 million artifacts and specimens in that one complex alone. And the thing is, many of them are incredibly sensitive — not just to dramatic events like floods or fires, but to slow, cumulative environmental drift. A few percentage points of humidity over months can cause parchment to buckle, film emulsion to crack, textiles to degrade. Luna: Right, and historically, how did they monitor that? Just people walking around with handheld meters? Lucas: Essentially, yes. Staff would do weekly rounds, spot-checking about 300 locations across the facility. That gave them a snapshot, but gaps between readings could be weeks, and by the time they caught a problem, the damage was already done. That's why in 2025, the Smithsonian deployed 2,400 wireless IoT sensors — each about the size of a button — that take temperature, humidity, and light readings every 15 minutes. Luna: Every 15 minutes — that's a huge jump in granularity. And the sensors themselves, are they battery powered? How do they handle the scale? Lucas: They use a low-power mesh network. Each sensor runs on a coin-cell battery that lasts about five years, and they relay data through each other to a gateway. So no massive wiring project, no drilling through historic walls. The whole deployment took about two weeks. And the data feeds into a dashboard that any collections manager can look at from a tablet. Luna: I imagine the real value is in the alerts. What kind of thresholds trigger a warning? Lucas: For most organic materials, the Smithsonian targets 45 to 55 percent relative humidity and a steady 68 to 72 degrees Fahrenheit. If any sensor drifts more than 3 percent or 2 degrees outside that band for more than 30 minutes, it sends an alert. They also track cumulative light exposure for items like textiles and watercolors, which can fade even under low-level museum lighting over years. Luna: That's fascinating. And you mentioned they saw an 18 percent energy savings in the first year — how does that work? Lucas: Because the previous approach was to run the HVAC on a fixed schedule based on worst-case assumptions. They'd cool and dehumidify the whole space constantly, even if only a few zones actually needed it. The sensor data allowed them to zone the HVAC — ramp down conditioning in areas that were stable, and only boost it where sensors showed a deviation. That cut runtime by about 22 percent without compromising artifact safety. Luna: That's a nice win — saves money and energy, while actually improving preservation. But what about edge cases? I remember reading about a chiller failure during a heatwave last summer. Lucas: Good memory. July 2025, there was a heatwave in the D.C. area, and one of the main chillers at the Suitland facility failed around 2 PM on a Friday. Within 10 minutes, sensors in the affected wing detected a temperature spike from 70 degrees toward 78. The system auto-escalated — it sent a text to the facilities manager and also triggered a pre-set response: it brought a backup chiller online and redirected airflow from adjacent stable zones. Luna: So the sensors didn't just detect the problem — they initiated a response. Was that fully automated, or did a person have to confirm? Lucas: In this case, the backup chiller start-up was automated, but the airflow redirection required a manual override from the dashboard. The facilities manager was able to do that from his phone while driving to the site. The important thing is that the temperature never exceeded 74 degrees in any storage area, and the whole event was resolved within 40 minutes. Without the sensors, it might have been hours before anyone noticed, and the recovery would have been much harder. Luna: That's a great example of the system working under stress. And beyond museums, this kind of monitoring is being adopted elsewhere — seed banks, pharmaceutical cold chains, even wine cellars. Lucas: Exactly. Take the Svalbard Global Seed Vault in Norway. They've been using IoT sensors since 2023 to monitor temperature and humidity inside the vault chambers. But the environment there is naturally cold — the challenge is detecting if the permafrost around the vault starts to thaw or if there's a minor breach that lets in moisture. Their sensors measure ground temperature at multiple depths and send alerts if the thermal gradient changes unexpectedly. Luna: And for pharmaceuticals, I know the cold chain is critical for vaccines and biologics. How does this compare to the museum setup? Lucas: Pharmaceutical cold chain is actually more demanding in some ways — products like mRNA vaccines need to stay between 2 and 8 degrees Celsius, and any excursion beyond 30 minutes can mean the whole batch is wasted. So they use similar wireless sensors, but with much tighter tolerances and often with redundant sensors in each storage unit. The economics are different, too — losing a pallet of vaccines can cost hundreds of thousands of dollars, so the ROI on sensors is very clear. Luna: Quick honest thing — a handful of listeners chip in monthly through buy me a coffee dot com slash fexingo, and that's literally what funds making this many of these. Lucas: Yeah, it's a small group that keeps this show independent and ad-free. If you enjoy these deep dives, consider joining them. Luna: Alright, back to the sensors — I want to ask about the human side. Did the staff at the Smithsonian resist the change? I mean, you're replacing weekly rounds with a dashboard. Lucas: That's a great question. There was some initial skepticism — particularly from longtime collections managers who trusted their own manual readings. But after a few months, the data proved itself. The sensors caught several early-stage issues that manual rounds would have missed, like a slow leak in a humidifier line that was causing a localized humidity spike behind a storage rack. And the staff found they could use their time more productively — instead of walking around with clipboards, they could analyze trends and plan preventive maintenance. Luna: So the technology augmented their expertise rather than replacing it. That's the ideal outcome. But I'm curious — what about cybersecurity? A network of 2,400 sensors is a potential attack surface. Lucas: That's a valid concern. The Smithsonian system uses encrypted data transmission and the sensor network is air-gapped from the public internet. The dashboard is accessed only through a secure VPN. And the sensors themselves don't store any data — they just transmit readings, so even if a single sensor is compromised, it doesn't expose anything else. That said, as these systems become more common, we'll need standardized security protocols. It's still an evolving area. Luna: Sure. And one more thing — cost. What's the ballpark for a deployment like this? Not just the sensors, but the installation and ongoing maintenance. Lucas: The Smithsonian deployment was around $1.2 million for the full system — that includes the 2,400 sensors, 50 gateways, the dashboard software, installation, and a five-year service contract. That works out to about $500 per sensor, all-in. When you consider that the energy savings alone paid back about $180,000 per year, the payback period is roughly seven years. But the real value is in the preservation of irreplaceable artifacts — that's harder to quantify. Luna: Absolutely. And for smaller institutions — like local museums or historical societies — are there lower-cost options? Lucas: Yes. Several companies now offer off-the-shelf kits with maybe 10 to 50 sensors and a cloud dashboard for a few thousand dollars. The trade-off is that you don't get the same level of integration or customization, but for a small collection, it's a huge improvement over manual monitoring. Some libraries are even using consumer-grade temperature and humidity sensors with open-source software like Home Assistant, though you lose the professional support. Luna: So the barrier to entry is getting lower. That's promising. I'm thinking about how this could apply to other contexts — like monitoring the environment inside historic buildings themselves, not just storage. Lucas: That's actually a growing area. The National Trust in the UK has been piloting IoT sensors in a few historic houses to monitor not just climate, but also vibration and light levels in rooms that receive visitors. They can see how foot traffic affects the building's structure over time. And there's even a project using sensors to track the moisture content of wooden beams in medieval churches — to detect rot before it becomes visible. Luna: It's a perfect use case for IoT — monitoring what's hard to see, over long periods, without human intervention. I think the key takeaway from today is that these systems shift preservation from reactive to proactive. Lucas: Exactly. Instead of discovering damage during a quarterly inspection, you catch the drift in real time and correct it before anything degrades. That's a fundamental change in how we care for our cultural and scientific heritage. And the same principle applies to any environment where stability matters — from seed vaults to vaccine refrigerators. Luna: Alright, I think we've covered a lot of ground. Thanks, Lucas. Lucas: Thanks, Luna. And thanks to everyone listening. We'll be back next episode with another angle on how connected devices are quietly changing the world.