Device and software costs are what everyone negotiates. Maintenance is the number that determines whether your business case holds at year five. Good vendors engineer for both.
This article is about remote IoT monitoring programs - sewer level monitoring, water network pressure, flood sensors, tanks, pump stations. The kind of infrastructure monitoring where devices go into the ground or into the field, and someone has to maintain them for a decade.
The full cost schedule: what your business case needs to account for
Every monitoring deployment carries the below cost structure. How much it costs you depends entirely on how well your vendor has engineered against it.
| Cost Driver | What It Means for Your Program |
|---|---|
| Battery replacement frequency | Every battery swap is a site visit. The difference between a 3-year cycle and a 10-year one adds up significantly. |
| Battery swap design | If a swap requires device removal, the labor cost dwarfs the battery cost. The design of the replacement process matters as much as the interval. |
| Install time | The difference between a 30-minute install and a 2-hour+ one compounds fast across a network. |
| Device reliability | A device that fails in a flooded pit or loses signal underground is a data gap with operational consequences. |
| Alarm noise | Every false alarm that sends a truck is a maintenance cost. |
| Maintenance overhead | Complexity at install becomes complexity at every service visit. It compounds as the network grows. |
| Platform and vendor fragmentation | Solving gaps by adding vendors costs more than preventing the gaps in the first place. |
The table above is what you are really evaluating when you choose a monitoring vendor. Not just the device, or the connectivity, or the software subscription but whether their entire system (hardware, battery design, installation model, platform) has been genuinely engineered to keep your maintenance burden down over the life of the program.
That is the most important question in any monitoring procurement. More important than unit price. More important than the software tier. Because maintenance cost doesn't show up on day one. It shows up at year three, year five, year seven - doubling or tripling what you thought the program would cost, long after the business case has been approved and the vendor relationship is locked in.
The business case that only models device price is only modelling the entry fee.
This isn't a criticism of how business cases get built, it's a reflection of what standard procurement frameworks are designed to compare. Unit cost is visible, auditable, and easy to benchmark across vendors. A 10-year maintenance model is harder to build and harder to defend without real data to underpin it.
The utilities that manage monitoring programs well close that gap deliberately. They ask different questions before procurement is finalised. They push vendors to model lifecycle cost, not just device price. And they look closely at the design decisions behind the hardware and the platform, because those decisions are what determine what the program costs to run at year three, year five, and year ten.
What these cost drivers look like in practice
Battery life: the most predictable cost driver, and the most underestimated
Battery replacement is the line item that surprises most utilities when they run the numbers properly. Not because it's hidden, everyone knows batteries need replacing, but because the scale of the difference between vendors rarely gets modelled at procurement.
Devices built to run for 10+ years on a single field-swappable battery represent a fundamentally different maintenance burden to those operating on a 3–5 year cycle that require device removal. Even at 500 devices, that difference is not marginal. We've modelled what it looks like - it's worth reading before any procurement decision is finalised.
See our dedicated piece on IoT battery life and total cost of ownership →
Install complexity: a one-time cost that keeps compounding
Installation cost is easy to underestimate because it gets calculated per device. At 10 sites it barely registers. At 500+, every additional minute per install is a meaningful project cost, and every layer of complexity becomes a constraint that reappears every time a device needs servicing.
What separates a fast, scalable install from an expensive one comes down to a handful of design decisions the vendor either made or didn't:
- No specialist tooling - any trained technician can do the work, not just the two people on your team who've done it before.
- No confined space entry requirement - faster access, fewer permits, and lower risk at every visit.
- Clip-on mounting systems - no structural modification means installs that can be completed and reversed without a contractor.
The software side matters just as much. A mobile app that walks a technician through configuration on site, like setting obstruction masking, checking signal strength, entering pipe depths and alarm thresholds, validating live data before they leave, removes the most common source of post-install problems. Everything confirmed on site, from a phone, before the manhole lid goes back on. No laptop, no return visit to fix a configuration error that wasn't caught at install.
The speed of install reduces deployment cost from day one. And because the same simplicity applies every time a device is serviced or swapped, it keeps reducing cost for the life of the program.
Device reliability: the cost that's difficult to budget for
Every monitoring program will encounter device issues at some point. The question is whether you're confident your vendor has done everything possible to make sure it doesn't happen often.
The more useful question to ask a vendor is where their devices have actually been deployed, for how long, and what the failure data looks like.
IP68 is a baseline, not a guarantee. It tells you a device can handle temporary submersion under controlled test conditions. What it doesn't tell you is how a device performs after weeks submerged in a live sewer network, exposed to hydrogen sulphide, pressure fluctuation, and physical shock.
When a device does fail, the cost is the truck roll to diagnose it, the crew time to re-deploy, and the data gap created. A vendor who has invested seriously in field-proven reliability is reducing that exposure - not just the replacement cost, but the operational consequence of the failure itself.
Alarm noise: the hidden crew cost
False alarms are a maintenance cost that almost never gets modelled at procurement, but for any utility running a large monitoring network, they add up fast. Every unnecessary callout is crew time, vehicle cost, and an hour or two that could have been spent elsewhere. A platform that gives your team the context to verify before they dispatch is a direct reduction in ongoing maintenance spend.
Scale: where weak design choices become expensive ones
Platform scalability is the constraint that tends to surface last and costs most to fix. Any monitoring solution handles 10 sites. The real test is at 500 or 2,000, or when you need the same platform to extend across sewer, flood, pressure, tank, and flow monitoring without adding new vendor relationships and new data silos.
The monitoring programs that scale well share a common characteristic: the vendor built for it from the start. Fleet management that works the same way at 20,000 devices as it does at 20. Deployment templates that ensure consistency across a large rollout without manual repetition. A platform that carries multiple use cases in a single view, across a single data environment.
The vendor questions that actually matter
The standard procurement question is which vendor offers the best device at the best price. It's a reasonable starting point. It's just not the question that determines what your monitoring program costs to run.
A vendor who has genuinely engineered for low total cost of ownership can answer every question below with specifics. The answers are in the design of the hardware, the architecture of the platform, and the maintenance model they stand behind.
| Before you commit — questions worth asking every vendor | |
|---|---|
| 1 | What is the realistic battery life in our field conditions (not lab conditions) and what does replacement require in terms of crew, access, and time? |
| 2 | What does install actually look like across a large rollout? What is the average time per site, and what specialist requirements does it carry? |
| 3 | How does the platform manage alarm noise? Can your team see what triggered an alert and verify whether a site visit is warranted before dispatching? |
| 4 | Can you show us real-world device performance in the specific conditions our network operates in? |
| 5 | How does the platform scale? What does managing 500 devices look like versus 50, and does complexity grow with it? |
| 6 | Can the same platform extend across multiple use cases (sewer, flood, pressure, tanks, flow) or does each one add a new vendor or new solution into the mix? |
| 7 | Will you help us model the 10-year cost of this program not just the upfront price? |
The distinction worth making before you sign anything
A monitoring program that looks affordable at procurement can become one of your most expensive infrastructure decisions if the cost structure isn't right. You won't feel it on day one. You'll feel it at year five, when the maintenance budget has quietly outgrown the original business case and scaling the program is harder than it should be.
The vendors who understand this design for the full cost of ownership. And that design is either saving you money every year the program runs, or costing you money you never planned to spend.
