An IoT platform is the software layer between your devices and your decisions. It receives readings from sensors, meters, controllers and cameras, gives every value a place and a time, and makes the data usable: as a dashboard, an alert or a direct answer to a question. Without a platform, every manufacturer brings its own app, and nobody sees the whole building.
Why do you need an IoT platform at all?
A single sensor is easy. It measures one thing and shows it in one app. The trouble starts with the tenth device from the third manufacturer. The heating runs on BACnet, the meters speak M-Bus, the new room sensors send over LoRaWAN. Each system has its own login, its own units and its own alarms.
An IoT platform brings these sources together. It answers questions that no single device can answer, such as "Why does store 07 use more energy than store 04?" or "Which rooms were too humid after last week's rain?". The value is not in the individual reading. It is in the comparison, the history and the context.
What are the layers of an IoT platform?
Most platforms follow the same structure from the field to the screen. The names differ, the jobs are the same. The table shows the four layers and how baiond covers each of them.
| Layer | What it does | How baiond covers it |
|---|---|---|
| Devices and sensors | Measure temperature, humidity, energy, people, fill levels and more | Wireless sensors, existing meters and building automation, cameras, satellite imagery and weather |
| Connectivity | Carries the readings from the field to the platform | The baiond Gateway reads LoRaWAN, M-Bus, Modbus, BACnet, KNX, DALI and MQTT and turns them into one data stream |
| Data platform | Stores, structures and contextualises every value | The Space System puts each reading in its place: site, building, floor, room, device |
| Applications and AI | Turns data into views, alerts and answers | AI Agents watch the data continuously, the Graph Builder builds charts without code |
Devices and sensors
The field layer is everything that measures. Many sites already have more data sources than they think: heat meters, energy meters, the building management system, access control. New sensors fill the gaps where nothing measures yet. A good platform counts what you already have instead of asking you to replace it.
Connectivity
Readings need a way to travel. Wireless standards such as LoRaWAN and NB-IoT reach basements, roofs and outdoor areas with little power. Wired building protocols such as BACnet, KNX or Modbus already run in most larger buildings. The connectivity layer translates all of them into one format, so the platform does not care which language a device speaks.
Data platform
This is the core. The platform stores time series, checks for gaps and knows which value belongs where. That last part matters more than it sounds. A temperature of 5.8 °C is meaningless on its own. It becomes useful once the platform knows it comes from the fish counter in store 04 and should stay below 4 °C.
Applications and AI
The top layer is what people actually use: dashboards, reports, alerts and, increasingly, AI that interprets the data. Here the term AIoT comes in. AIoT combines the Internet of Things with artificial intelligence. An AIoT platform does not stop at showing numbers. It explains them, compares them with history and weather, and recommends what to do next.
Which functions matter in daily operations?
Feature lists are long. In practice, a handful of functions decide whether a platform gets used after the first month.
- One view across all systems. Sensors, meters and building automation in one place, with one login and consistent units.
- A clear spatial model. You want to ask about "floor 3" or "all stores in Vienna", not about device IDs.
- Alerts with a reason. A message that says what is wrong, where, and why it matters, instead of a red dot.
- Early warnings. Predictive Maintenance learns how a machine normally behaves and flags drift early, often before it fails.
- Context from outside. Satellite and weather data explain readings: a wet wall after three days of rain means something else than in a dry week.
- Self-service views. Facility teams build their own charts without waiting for IT.
- Plain-language answers. You ask a question and get the reading, the likely cause and the next step.
What does data orchestration mean?
Buildings, sites and landscapes produce data in many formats: sensors, meters, controllers, cameras, satellites and weather services. Data orchestration means bringing all of it together, giving every value its place and time, and making it usable in one system. We call this data orchestration for the physical world.
The difference to a classic IoT platform is the range of sources. Many platforms focus on their own sensors. An orchestration platform treats the existing building automation, an external weather service and a camera count as equal sources. That is what makes cross-cutting questions possible.
How do you choose an IoT platform?
The right platform depends on your sites, your existing systems and your team. These criteria help you compare offers.
Does it work with what you already have?
Ask which protocols the platform reads directly. If your building runs on BACnet or KNX, the platform should connect to it without replacing the automation. Retrofitting existing buildings is the normal case, not the exception.
Can it grow from one site to many?
Start with one building, one store or one problem. Check how the platform handles the second and the twentieth site: are spaces, users and alerts easy to copy and adapt? Our plans are built around exactly this step from pilot to portfolio.
Who can use it?
A platform that only engineers understand stays with engineers. Look for views a store manager or a facility lead can read in seconds, and for a way to ask questions without learning a query language.
Where is the data stored, and who owns it?
Clarify hosting, export and ownership before you sign. At baiond, your data stays yours, it is hosted in Europe, and you can export your readings at any time. Camera analysis keeps only counts, never images.
What happens when a sensor is missing?
No catalogue covers every case. Ask whether the vendor can build or integrate a specific sensor. baiond develops custom-built sensors when a measurement is not available off the shelf.
What does an IoT platform look like in practice?
Take a supermarket chain. Every store has fridges, freezers, a heating system and energy meters from different manufacturers. Without a platform, the store manager checks displays by hand and learns about a failed compressor from the morning shift.
With a platform, every fridge reports its temperature, the gateway connects the existing meters, and the spatial model knows which counter is in which store. The AI agent notices that one compressor runs longer every day and warns early, before it is likely to fail. The manager asks "Is the fish counter in range?" and gets the reading, the likely cause and the next step. The supermarket page shows this scenario in detail.
Frequently asked questions
What is the difference between IoT and AIoT?
IoT connects devices and collects their data. AIoT adds artificial intelligence that interprets this data. Instead of raw numbers, you get explanations, early warnings and recommendations.
Do I have to replace my building automation?
Usually not. A gateway reads most existing systems over their own protocols, such as BACnet, KNX, Modbus or M-Bus. Your automation keeps running as it is, and the platform adds the view across all of it.
How long does it take to get started?
That depends on the number of sites and the systems already in place. Most projects start with one site and one clear question, then grow from there. We plan the first step with you in a consultation.
Is an IoT platform only for large companies?
No. A single building, a church or a sports facility benefits just as much as a chain with hundreds of sites. What matters is a clear question the data should answer.
Want to know what an IoT platform could answer for your sites? Request a consultation.
