Last year, we were of the opinion that a "somewhat" standard Internet of Things (IoT) solution stack remains a work in progress. Since then we have explored various components and it is clear that cloud-based platform and infrastructure services will play a key role in the realization of end-to-end IoT solutions. However, we expect hybrid middleware paradigms to be the norm for enterprise IoT initiatives.
Despite many flavors, IoT middleware will follow a cloud-centric evolution
An IoT middleware layer is a combination of integration platforms and capabilities providing the glue for connecting devices, users, and applications/services in a "cyber-physical" world. This combination can include an enterprise service bus (ESB)/service-oriented architecture (SOA) infrastructure, which is a good option for developing a middleware layer between devices and back-end applications/services. Moreover, an integration iPaaS (iPaaS) solution can be used for ingesting IoT data into business intelligence and analytics tools, and for integration with on-premise and software-as-a-service (SaaS) applications. It is practically possible to develop a purely/largely cloud-based IoT solution stack capable of supporting various integration, storage, analytics, and security requirements.
From a broader perspective, different flavors of IoT platform include
- application development and integration-centric IoT platforms
- device connectivity and communications-centric IoT platforms
- data storage and management-centric IoT platforms
- analytics and reporting-centric IoT platforms.
Some IoT platforms offer a wider range of capabilities and are well suited for more generic IoT initiatives. The evolution of cloud-based IoT platforms (which can be termed "IoT PaaS" in some cases) is a natural step considering the rise and initial success of adjoining PaaS sub-segments, especially iPaaS, mobile backend-as-a-service (MBaaS) or relatively comprehensive mobile platform-as-a-service (mPaaS), application platform-as-a-service (aPaaS), and data platform-as-a-service (dPaaS) or data-centric PaaS.
Cloud-based infrastructure and platform services offer elastic scalability and can support high-throughput event processing, which are key requirements for realizing value from IoT. In the context of IoT platforms, what several vendors are offering is a set of cloud-based services that can cater for specific IoT functions, however, these services may or may not be sufficient for developing an end-to-end IoT solution. For this reason, we expect a significant number of enterprise IoT initiatives to follow hybrid middleware paradigms involving the use of on-premise middleware and cloud-based IoT and/or integration platforms.
In the near-term, we expect significant consolidation in the IoT platform vendor landscape, with major vendors and service providers acquiring smaller IoT platform vendors. This will be followed by product integration to develop relatively comprehensive IoT solutions. In this context, PTC's acquisitions of ThingWorx and Axeda, further complemented by acquisitions of ColdLight for big data analytics and Vuforia for augmented reality, are good examples. Given the range and complexity of IT requirements, an evolving IoT middleware layer will retain its cloud-centric nature, with on-premise middleware used for meeting specific integration needs.
Saurabh Sharma, Senior Analyst, Software – Infrastructure Solutions