Feeling overwhelmed by the architecture of IoT? Sensors, actuators, communication protocols — it can feel like a Herculean task, especially for newbies. You might be wondering, “Where do I even start?”
That’s what WebbyLab is here for. We’ll share a guide to the Internet of Things (IoT) architecture, backed by our 5-year expertise. If you’re tired of scrolling through endless guides without finding clear answers, consider this article your ultimate handbook.
Before designing an IoT product, it’s critical to understand how devices will interact with it. It’s necessary to decide on the communication protocols the platform should provide or include the possibility of expanding communication protocols within the IoT architecture.
Kostiantyn Oliynyk, Head of IoT at WebbyLab
As experts in various IoT niches like HVAC systems, consumer electronics, access control solutions, and smart homes, we’ve tackled different types of IoT architecture. We’ve successfully implemented them in projects like the Propuskator ACS and 2Smart Standalone automation platform, along with many others.
The industries WebbyLab works with
Drawing on our profound experience as an IoT development company, we’ll break down the IoT architecture, covering:
The 2Smart Standalone automation platform developed by WebbyLab
The four stages of the IoT architecture
While all IoT solutions are unique, their architecture can typically be broken down into four fundamental layers. Let’s explore each in greater detail.
This layer is the foundation of an IoT system. It’s where physical environment data is captured.
The perception layer comprises sensors responsible for measuring everything from temperature and humidity to motion and light. Actuators, another part of this layer, are used to translate digital signals into physical actions. For example, turn on the light or the air conditioning system in response to the certain data obtained.
The IoT device architecture essentially bridges the physical and digital worlds, collecting raw data.
The IoT network architecture layer enables the connection and communication between IoT devices and the platform. It’s possible owing to communication protocols like Wi-Fi, Bluetooth, MQTT, HTTP, Zigbee, and cellular networks. To effectively implement these protocols, businesses often rely on custom IoT software development services.
From WebbyLab’s experience, MQTT and HTTP are the most common network layer protocols.
Kostiantyn Oliynyk
Head of IoT at WebbylabHelped dozens of startups create effective IOT products.
Get a ConsultationRaw data collected by IoT hardware is just the beginning. The IoT data architecture layer is where the gathered information gets filtered, aggregated, processed, and interpreted.
This layer involves using various tools and techs for data analysis. Those include:
The above tools let you identify trends, predict future events, or trigger specific actions based on the insights obtained from data.
The IoT platforms typically provide real-time or historical data analytics capabilities as well. This functionality helps users gain insights into their IoT devices and data for further device/system performance optimizations.
This is essentially the IoT software architecture layer that interacts with end-users directly. It typically provides interfaces and features that let users monitor data, manage devices, handle firmware updates, parameter adjustments, and diagnostics.
The IoT software platform architecture layer may include:
WebbyLab`s IoT software development services can help you build custom robust solutions!
An IoT platform is the brain of your entire IoT ecosystem. It collects, processes, analyzes, and visualizes data from your connected devices. But how does it work, and what does it consist of?
See the common IoT platform components that WebbyLab experts typically use on the following IoT architecture diagram:
The diagram of a common IoT architecture
Now, let’s discuss each IoT platform architecture component in greater detail.
The load balancer is the entry point for any web application. It ensures a smooth traffic flow throughout the platform and prevents servers from becoming overloaded.
This part is especially critical for the IoT platform because it connects thousands, tens or hundreds of thousands, even millions of devices simultaneously. Continuous connection is a key consideration here.
WAF is a specialized type of network hardware or software designed to protect web applications from various types of attacks.
MQTT is a transport protocol used for both connecting devices to the platform and transferring data from the devices to the end-user or platform.
Services is a set of functionalities that make up the IoT platform. These include typical front-end/back-end services as well as any additional services that expand the platform’s capabilities or functionality.
Some of the most helpful services include the following ones:
It’s better to collect historical data from the very outset of the platform, even if there are no plans for analysis. When it becomes necessary to conduct an analysis, there will already be a sufficient sample of historical device data.
Kostiantyn Oliynyk, Head of IoT at WebbyLab
A monitoring system is a set of services for monitoring the system’s status. Given the continuous connection of a large number of devices, configuring failure alerts is critical to ensuring system stability.
The IoT platform monitoring system
While not every IoT architecture relies on MQTT, WebbyLab frequently leverages this protocol for its lightweight design, efficiency, and reliability. Let’s consider several examples of MQTT-based projects implemented by our experts.
The IoT architecture of the Propuskator ACS
If we compare our access control system, Propuskator, with the IoT architecture diagram above, they are nearly identical. Our Propuskator devices connect to the platform via an MQTT broker and use HTTP to synchronize access schedules/rules and access logs.
The only distinction is the absence of historical device data collection and storage, as ACS doesn’t require it. It serves a different business task.
The IoT architecture of the 2Smart Standalone automation platform
Let’s now look at our other project with a slightly different architecture. 2Smart Standalone isn’t inherently an IoT platform but a BAS (Building Automation System). Yet, it can also be considered an IoT device management solution.
The primary distinction lies in its ability to integrate devices from various manufacturers and communication protocols through bridges. The bridge is a service that internally converts one protocol into another, facilitating connectivity between the device and the platform.
For example, Modbus devices communicate with each other via the RS485 protocol using cables. The bridge converts this communication from cables to MQTT and vice versa. Consequently, the platform’s end-user can send a command through the platform; the bridge will receive it, convert it into RS485, and transmit it to the device.
Let’s recap!
While the above diagrams are different, their core element remains the same — the MQTT broker (connectivity). Thus, the connectivity component can be considered the cornerstone of the IoT platform architecture.
Now that we’ve explored the IoT architecture our team usually uses, let’s consider an alternative approach. Here’s an example of the architecture based on Google Cloud Platform that doesn’t rely on an MQTT broker:
An IoT architecture without MQTT
In this IoT network architecture setup, devices use the back-end’s REST API for sending and receiving necessary information.
The distinctive feature of these types of IoT architecture is the lack of real-time device management. Users can input certain settings, which the device will receive upon contacting the back-end, but only when it initiates a request.
In this example, devices are typically in the sleep mode to conserve battery power. They wake up according to a schedule, communicate with the back-end to receive parameters for their tasks, send reports, and then return to sleep mode.
This GCP-based architecture uses various ready-made tools from Google Cloud. This greatly facilitates development but increases monthly infrastructure costs.
GCP is an excellent tool and definitely worth the money for IoT solutions. However, it should be used in cases where there is a real prospect of production and high loads that GCP can easily handle. Startups may end up spending a large amount just on the infrastructure even before reaching the production stage.
Kostiantyn Oliynyk, Head of IoT at WebbyLab
The application we developed consists of the following:
Hence, users interact with the system via web browsers or a mobile application, administrators access it through the web admin panel, while devices communicate directly with the back-end via the provided API. To ensure seamless integration of all components and optimal system performance, many companies engage custom IoT consulting to design and implement tailored architectures.
Let’s examine the components of this IoT cloud architecture in greater detail. Here’s the breakdown of each Google Cloud service you can see in the diagram above:
And here’s an in-depth explanation of some of the Google Cloud services we leverage:
The WebbyLab team used bespoke services as part of this IoT system architecture based on GCP. Here’s what we added:
All services deployed on Cloud Run are secured behind load balancers, which are further protected by Cloud Armor and its policies.
Wonder how to put those IoT architecture ideas to good use? Check out the top 10 IoT use cases.
Building a secure and scalable IoT architecture is key to your project’s long-term success. Here’s how to make your system stronger:
A decent IoT security architecture ensures protection layer by layer. The methods you may leverage include:
Encryption makes your data unreadable to unauthorized parties. Implement it to protect your data throughout its entire journey, from device to cloud and back.
Plus, don’t forget about access control measures. These make sure that only authorized users and devices can access/read/modify specific data within your Internet Things system.
If you aim to develop a future-proof IoT architecture, leaving room for scaling is mandatory. To do so:
WebbyLab has been delivering robust IoT architectures for over five years. During this time, our expert team has tackled 10+ projects and developed three IoT products of our own, including 2Smart Standalone and Propuskator.
A mobile app of the Propuskator ACS
We can handle a range of services to facilitate your IoT platform architecture:
But less words, more action. See how we implement successful IoT solutions in our portfolio.
IoT architecture might seem complex, but with the right foundation and partner, you can achieve success. Our article has just covered the basics — the possible IoT applications are truly limitless.
Have further questions on the architecture of IoT? Contact our experienced specialists at WebbyLab to turn your project idea into a robust, viable ecosystem.
Written by:
Head of IoT at Webbylab
With a robust academic background in Telecommunication Systems Engineering, I apply my knowledge to lead innovations in the IoT domain. Starting as the first team member in the newly formed IoT department at WebbyLab, I've spearheaded its growth, fostering the expansion into embedded and hardware development alongside our core software projects. My dedication lies in pushing the boundaries of IoT technology, fostering a culture of innovation and excellence that profoundly impacts our clients' operational success.
In IoT end-to-end architecture, a gateway is an intermediary between the devices and the cloud. It facilitates bidirectional communication and seamless data flow.
The typical stages of building an Azure IoT reference architecture or any other setup are:
Sensors are used at the very first stage of the IoT architecture. They are installed throughout the facility to collect data on parameters like temperature, motion, light, and beyond.
The cloud lets you store, process, and analyze data from IoT devices. It’s not only scalable and secure but also enables remote monitoring and management of devices. On top of that, it can be enhanced with machine learning algorithms and other tools for data analysis.
The data processing and analysis layer is responsible for aggregating and analyzing data in industrial IoT architecture or any other solution. It often uses advanced analytics and machine learning algorithms to turn raw data into valuable insights.
2025 WEBBYLAB LLC. All rights reserved.
Get a consultation with at Webbylab
