IoT platforms emerged early in the history of IoT, arguably too early. At the time, the industry was driven by momentum and the urgency to provide ready-to-use solutions.

Over the past decade, however, many of these platforms have been discontinued, transformed, or fundamentally redesigned. This evolution is not accidental; it reflects a deeper structural issue in how the first generation of IoT platforms was conceived.

A flawed starting point: visualization over operations

The earliest IoT platforms were built by teams with limited real-world experience in large-scale IoT deployments. As a result, they relied on familiar paradigms and existing technological building blocks. At that time, users were primarily looking for sensor data visualization, and the perceived value of IoT platforms was strongly associated with business intelligence capabilities.

Cloud providers such as AWS, Microsoft, and Google quickly delivered solutions that essentially functioned as data pipelines: device ingestion through brokers, chained storage layers, and visualization via existing BI tooling.

This architecture borrowed heavily from big data concepts : massive incoming datasets, scalable storage, rapid visualization, yet it failed to address the true nature of IoT system management.

In practice, most IoT fleets consist of thousands to hundreds of thousands of devices, each transmitting very small amounts of data. The hyperscale infrastructure behind early IoT platforms therefore produced:

• High operational costs
• Excessive scalability relative to real needs
• Limited support for real-world fleet management processes

Most importantly, these platforms did not solve the core problem of IoT: operational lifecycle management.

Ten years later: IoT platforms are not BI tools, they are ERPs

Ten years later, field experience has made one conclusion unavoidable. The core of an IoT system is not visualization. An IoT platform is not, in essence, a BI tool. It behaves much more like an enterprise resource planning system.

Managing a fleet of connected devices is ultimately similar to managing any commercial product lifecycle. There are manufacturing stages, storage and logistics, delivery and commissioning, maintenance operations, updates, and end-of-life handling. What makes IoT fundamentally different is the permanent technical complexity embedded in that lifecycle. Devices require firmware evolution, connectivity renegotiation, identity provisioning, authentication mechanisms, and continuous interaction with users. Regulatory pressure, already visible in existing frameworks (RED) and becoming unavoidable with the Cyber Resilience Act, adds another layer of obligation around security, updates, and data protection.

Configurability over out-of-the-box promises

Because early IoT platforms were not designed around these operational realities, companies had no choice but to compensate. Engineering teams progressively wrote enormous amounts of custom code to manage device lifecycles, permissions, user relationships, secure data handling, metadata orchestration, and heterogeneous communication channels. Step by step, organizations rebuilt internally what the platforms should have provided from the beginning: structured operational governance. In other words, they rebuilt an ERP.

This is why the real transformation happening today is not an incremental improvement of IoT platforms but a conceptual shift. The platform, as originally imagined, is fading away. What is emerging instead is the IoT ERP ; software whose primary purpose is not to display data but to organize, secure, and operate the entire lifecycle of connected products.

In such a system, visualization becomes secondary. The real value lies in lifecycle control, identity and access governance, secure ingestion and decoding of data, metadata management for fleet reliability, structured interaction with users, and the ability to guarantee compliance while orchestrating updates across deployed devices. Deploying IoT at scale requires the same level of process maturity as finance, logistics, or supply chain management inside traditional enterprises. Just as companies rely on systems such as SAP or Oracle to structure their internal operations, connected fleets now require ERP-grade software to structure their operational reality.

A true IoT ERP cannot be delivered as a simple out-of-the-box product. Each organization carries its own processes, constraints, device behaviors, and regulatory exposure. The software must therefore be deeply configurable and extensible, combining domain expertise in IoT with an understanding of enterprise-specific workflows. This need for adaptation is not a weakness; it is precisely what defines ERP-class systems

Emerging solutions and market direction

The market is already beginning to reflect this transition. New solutions are being designed around governance and lifecycle control rather than dashboards alone. Some are still emerging, others are already in use, but they all point in the same direction: toward structured operational management of connected fleets rather than pure data visualization.

Engineering the Next Generation of IoT Operations

From a personal standpoint, this vision is not purely theoretical. I built my first IoT platform more than a decade ago, and because I was designing the devices themselves while also operating the associated services, the challenges of manufacturing workflows, maintenance processes, and lifecycle governance became apparent very early. That experience led me, about two years ago, to begin developing a new, far more generic platform designed explicitly around an ERP-like enterprise usage model. While the platform is not yet fully ready for broad public exposure, elements of it are already accessible on my GitHub. It remains largely open source, complemented by a paid layer for advanced capabilities, and is already running in several of my own projects where I continue to invest a significant portion of my time and energy.

This ongoing work also explains my reduced publication rhythm recently, as the construction of this platform represents, in my view, a foundational step toward what is truly required to operate and sustain IoT fleets in the years ahead.

Regulation as an accelerator

Regulation is accelerating this shift. The Cyber Resilience Act makes lifecycle security and update governance mandatory, while European initiatives around industrial data sharing impose additional control requirements. In this environment, traditional IoT platforms are no longer sufficient. Only ERP-grade systems can provide the level of control, traceability, and compliance that the coming decade will demand.

The first era of IoT was shaped by data collection and dashboards. The next era will be defined by processes, governance, and lifecycle mastery. This is not a marketing evolution but a structural one, driven by real deployments and real constraints.

Conclusion

What is certain today is that the landscape is moving quickly. The questions surrounding IoT governance are no longer limited to companies traditionally seen as managing connected fleets, because in practice almost every physical product is becoming, or will soon become, a connected one.

For any organization designing devices or building services around them, it is therefore essential to reflect, early in the conception phase, on how a complete IoT operational framework will be deployed. Not merely from a technical connectivity standpoint but across the entire lifecycle and process chain.

Having been deeply involved in these challenges for many years, I am always open to exchanging with teams facing similar questions. Whether to explore concrete solutions or simply to share field experience, these conversations are invaluable, and I would be genuinely glad to continue them with anyone engaged in this transformation.