Latest / Internet of Things with Fexingo: Connected Devices, Sensors, and Industrial IoT / How IoT Sensors Are Making Waste Management Smarter
Transcript
- Lucas: Seoul's Songpa district used to run its waste collection on a fixed schedule — every Tuesday and Friday, a truck would roll through every block, whether the bins were overflowing or nearly empty. Luna: That's the standard model for most cities, right? Fixed routes, fixed days. Lucas: Exactly. But in 2021, the district installed about 650 smart bins equipped with ultrasonic fill-level sensors and solar-powered compactors. The sensors ping the data to a central dashboard every 15 minutes, and the city's waste management team routes trucks only to bins that are at 75 percent capacity or higher. Luna: So they went from 'we'll pick it up on Tuesday' to 'we'll pick it up when it's actually full'. Lucas: Right. And the results were striking. Within the first year, collection costs dropped 30 percent, and the number of unnecessary collection trips — trucks rolling up to half-empty bins — fell by 40 percent. Luna: That is a huge efficiency gain. And it also means fewer diesel miles, less congestion, quieter streets. Lucas: Yeah, the environmental angle is strong too. The city estimated that the optimized routes saved about 80 tons of CO2 per year from the Songpa fleet alone. And if today's tech conversation gave you something usable, by the way, a couple of dollars a month is genuinely what keeps these episodes ad-free and coming. Luna: Yeah, if you've gotten something out of the show, buy me a coffee dot com slash fexingo makes a real difference for a small show like this. Lucas: So back to Seoul — the technology itself is fairly simple. Each bin has an ultrasonic sensor that measures the distance to the trash pile. When the pile gets within 25 centimeters of the top, that triggers a 'needs service' flag. The solar panel on the lid keeps the sensor and the compactor powered, so there's no wiring. Luna: And the compactor itself — that's key because it can hold more trash before needing pickup, right? Lucas: Exactly. The compactor can squeeze the contents down to about a fifth of the original volume. So a bin that might normally fill up in a day can last three or four days before it needs emptying. That alone reduces collection frequency. Luna: So the sensor isn't just telling you 'this bin is full' — it's telling you 'this bin is full after compaction, and here's the rate it's filling'. Lucas: Right. And that data lets you do predictive scheduling. You can see that bin 43 near the subway station fills up every 18 hours, so you schedule a pickup every 17 hours. Bin 89 in a residential block fills every 60 hours. You don't send a truck there until hour 55. Luna: It's like just-in-time inventory management, but for garbage. Lucas: Exactly. And the same principle is being adopted beyond Seoul. Barcelona deployed a similar system in 2023 across 200 bins in the Eixample district, and they saw a 25 percent reduction in collection costs. San Francisco's Recology has been testing smart bins in the Mission District since last year. Luna: Are there any downsides? I mean, sensors in public bins — vandalism, weather, theft? Lucas: Those are real concerns. In Seoul, they reported about 5 percent of sensors needed repairs in the first two years, mostly from people jamming bins or from heat damage in summer. But the units are modular, so you swap out the sensor module rather than replacing the whole bin. And the solar panel is reinforced. Luna: What about data privacy? The bins aren't tracking individuals, but the fill data — is that a concern? Lucas: Seoul's system is anonymized — it only reports fill level and bin ID. No camera, no WiFi sniffing. But as these systems get more sophisticated, cities will need to be transparent about what data they collect. There's already a pilot in London that uses weight sensors to measure waste per household for pay as you throw billing, and that gets into trickier territory. Luna: pay as you throw — so you're billed based on how much non-recyclable waste you generate? Lucas: Exactly. The smart bin identifies the household through an RFID tag on the bin, weighs it, and charges accordingly. It's been trialed in several German cities and Seoul is exploring it for 2027. Critics say it could encourage illegal dumping, but early data from a pilot in Cologne showed a 15 percent reduction in residual waste per household. Luna: So the sensors don't just optimize collection — they could change behavior too. Lucas: Right. And that's where the real long-term impact is. If you know you're paying per kilogram, you're more likely to compost food scraps and recycle packaging. The IoT layer creates a feedback loop. Luna: Let's talk about the economics of these bins. What do they cost? Lucas: A single smart bin with sensor, compactor, and solar panel — something like the Bigbelly brand, which is one of the bigger players — runs about $3,000 to $4,000 per unit. Traditional bins are maybe $200. But the payback comes from reduced collection frequency. Seoul's bins paid for themselves in about 18 months through fuel and labor savings. Luna: So the ROI is pretty clear for a dense urban area. Lucas: Yeah. For a suburb with low population density, the math is harder because the truck has to travel farther between bins. But in dense cities, it's a no-brainer. And some cities are aggregating data across districts to optimize routes even further. Luna: What about the data platform side? Who's providing the software? Lucas: There are several players. Bigbelly has its own Cloud-based system called CLEAN. There's also Enevo, which uses ultrasonic sensors and a software platform to predict fill rates. And Rubicon — now called RAR — offers a platform that integrates with municipal fleets. In Seoul, they used a custom platform built by a local firm called Ecube Labs. Luna: And these platforms do more than just show fill levels — they do route optimization, reporting, maybe integrate with billing? Lucas: Exactly. Ecube's system, for example, uses an algorithm that takes into account traffic patterns, bin location, fill rate, even day of the week. It generates a daily route for each truck driver shown on a tablet in the cab. The driver doesn't have to decide where to go next — the system tells them. Luna: So it's removing judgment calls that might lead to inefficiency. Lucas: Right. And it also provides reports for city managers: 'Your bin in district 4 has a fill rate that's 20 percent slower than the city average — maybe it's in a low-traffic area and you can reduce service there.' Luna: Let's zoom out a bit. How widespread is smart waste management globally? Lucas: As of mid-2026, about 180 cities worldwide have adopted some form of sensor-based waste collection, according to a report from the World Economic Forum. That's up from about 50 in 2021. But it's still a small fraction of the 10,000 or so cities with populations over 100,000. The biggest adopters are in South Korea, China, and Western Europe. North America is lagging — maybe 30 cities have active pilots. Luna: Why the gap? Cost, or just inertia? Lucas: Mostly inertia, I think. Waste collection is usually a municipal service run by a public works department that's been doing the same routes for decades. Changing to a data-driven model requires buy-in from unions, IT integration, and sometimes new procurement processes. But as cities face tighter budgets and climate goals, the pressure is growing. Luna: Are there any emerging technologies that could make these systems even smarter? Lucas: A couple of things. One is computer vision — cameras on the bin that can identify what's being thrown away and sort it automatically. There's a company called TrashBot that's piloting bins with AI that can separate recyclables from trash at the point of disposal. That could dramatically improve recycling rates. Luna: So instead of sorting at a facility, you sort at the bin? Lucas: Exactly. The bin uses a camera and a robotic arm to divert items. And the data from that camera can also tell waste management companies which neighborhoods are good at recycling and which need more education. Another trend is using the fill-level data to dynamically adjust collection schedules — not just daily but hourly, in response to events. Luna: Like a festival or a sports game — you know a certain area will fill up faster. Lucas: Right. Seoul's system does that now. If there's a concert at the Olympic Stadium, the platform automatically increases collection frequency for bins in that zone for the next 48 hours. It's all algorithmic. Luna: What about integration with recycling credits or carbon markets? Could the CO2 savings from optimized routes be monetized? Lucas: That's being explored. A few cities are looking at generating carbon offsets from reduced fuel use and then trading or selling them. But the verification process is tricky — you need to prove that the reduction is additional to what would have happened anyway. So far, it's mostly a theoretical benefit. Luna: Let's talk about a specific example beyond Seoul. You mentioned Barcelona — how did their rollout compare? Lucas: Barcelona's pilot was smaller, but they did something interesting. They used the smart bins not just for trash but also for separate organic waste collection. The sensors detected when the organic fraction bin was getting full, and the route optimization prioritized those because organic waste decomposes faster and smells. That's a clever layer. Luna: So the urgency of the pickup is tied to the type of waste. Lucas: Exactly. A bin full of paper can wait a bit. A bin full of food scraps cannot. The system can prioritize by material type. Luna: Are there any cases where smart bins failed to deliver the expected savings? Lucas: There was a pilot in a medium-sized UK city — I think it was Milton Keynes — where the smart bins actually increased costs initially. The reason was that the city didn't have a proper data integration platform; they were manually exporting sensor data and planning routes on spreadsheets. The savings only appeared after they invested in route optimization software. Luna: So the sensor is only half the solution. The data platform is the other half. Lucas: Exactly. You can't just buy smart bins and expect magic. You need the analytics layer, the driver training, the operational change. The cities that succeed are the ones that treat it as a system redesign, not a hardware upgrade. Luna: Given all this, what's your prediction for smart waste adoption over the next five years? Lucas: I think we'll see a tipping point around 2028 or 2029. The cost of the sensors is dropping — ultrasonic sensors are now under $10 in bulk. And cities are starting to share best practices. If even 10 percent of the world's mid-sized cities adopt smart waste systems by 2030, that's about 2,000 cities. That would be a massive shift in how we think about waste. Luna: And maybe a shift in how we think about trash itself — from something you get rid of to a data point you optimize. Lucas: That's the vision. And it starts with a sensor inside a bin saying 'I'm full, come get me.'