© All rights reserved, Alliance for Internet of Things Innovation (AIOTI) 2020
The most essential ones are:
• Stakeholders. There is a large variety of potential stakeholders with a wide range of roles that shape
the way each of them can be considered in the IoT system. Moreover, none of them can be ignored.
• Privacy. In the case of IoT systems that deal with critical data in critical applications (e.g., e-Health,
Intelligent Transport, Food, Industrial systems), privacy becomes a make or break property.
• Interoperability. There are very strong interoperability requirements because of the need to provide
seamless interoperability across many different systems, sub-systems, devices, etc.
• Security. As an essential enabling property for Trust, security is a key feature of all IoT systems and
needs to be dealt with in a global manner. One key challenge is that it is involving a variety of users
in a variety of Use Cases.
• Technologies. By nature, all IoT systems have to integrate potentially very diverse technologies, very
often for the same purpose (with a risk of overlap). The balance between proprietary and
standardised solutions has to be carefully managed, with a lot of potential implications on the
choice of the supporting platforms.
• Deployment. A key aspect of IoT systems is that they emerge at the very same time where Cloud
Computing and Edge Computing have become mainstream technologies. All IoT systems have to
deal with the need to support both Cloud-based and Edge-based deployments with the associated
challenges of management of data, etc.
• Legacy. Many IoT systems have to deal with legacy (e.g., existing connectivity, back-end ERP systems).
The challenge is to deal with these requirements without compromising the “IoT centric” approach.
A drawback of many current approaches to system development is a focus on the technical solutions, which
may lead to suboptimal or even ineffective systems. In the case of IoT systems, a very large variety of potential
stakeholders are involved, each coming with specific – and potentially conflicting – requirements and
expectations. Their elicitation requires that the precise definition of roles that can be related to in the analysis
of the requirements, of the Use Cases, etc. Examples of such roles to be characterised and analysed are:
System Designer, System Developer, System Deployer, End-user, Device Manufacturer.
In order to better achieve interoperability, many elements (e.g., vocabularies, definitions, models) have to be
defined, agreed and shared by the IoT stakeholders. This can ensure a common understanding across them of
the concepts used for the IoT system definition. They also are a preamble to standardisation. Moreover, given
the need to be able to deal with a great variety of IoT systems architectures, it is also necessary to adopt
Reference Architectures, in particular Functional Architectures such as the AIOTI High-Level Architecture.
A very large number of IoT platforms have been developed with the initial purpose of ensuring that a device
could interact with other devices or equipment, providing connectivity from point-to-point to more universal.
Standard Development Organisations (SDOs) and Standard Setting Organisations (SSOs) have developed a
number of approaches that focused on interoperability, initially at the network level and now well beyond.
Many standards have been defined with the possibility to serve as a basis for the development of platforms
that – in the best case - deal with interoperability in a generic manner, across a variety of business sectors, with
a variety of possible implementations. Such "standardised platforms" are relying on reference architectures,
interoperability stacks addressing different layers, generic protocol adaptors, etc.