The Australian IoE community has kicked of a plan to develop an IoE reference architecture that will provide a practical framework for the implementation of IoE systems, and it has already garnered support for the idea from around the world.
The planned architecture is an initiative of Peter Klement, a business technology consultant who, along with Glenn Vassallo, founded the Internet of Everything Sydney meetup group earlier this year.
Klement told IoTAustralia, that, while there were a number of IoT reference architectures in the public domain, they were mostly very generic, and very high level. “My experience is that they come from an ivory tower mentality, people trying to come up with a silver bullet that works for every industry and every use case,” he said. “That is one of the reason when many reference architectures have a bad reputation: they are very academic, very theoretical.”
He said the intent of the Australian initiative was to leverage existing reference architectures to create one that could be implemented in practice, choose products to fit it and then, in phase two, create some use cases using technology from vendors.
“What use cases we develop will be opportunity driven. Some industries are more interesting than others. I am hoping that people will bring problems or use cases to the table. Alternatively, we can reach out globally and see what is out there and reengineer it accordingly.”
Klement said the ideal outcome for the project would be for the reference architecture to be used in the implementation of End-To-End IoT solution in the real world.
Global participation
He said initial responses to the idea had been very positive. He has approached others, including the Australasian Architecture Meetup Group, seeking participation and has had offers to participate, voluntarily, from about 20 people. “We have people in Singapore, India, Brazil and Europe already interested to participate. It will be a global open source project, but it started and will be managed from Australia,” Klement said. He added: “The Sydney IoE community is growing into a regional IoE community network and we expect that the community network will eventually be global, like the IoE reference architecture project.”
He said the response to the initiative indicated that the community saw value in the planned architecture. “I got feedback from one person who works for a big telco who said: ‘If I had had something like this in my last project it would have been very helpful.’ If you don’t use reference models and you don’t use pieces of code that are out there, you have to develop both from scratch for a customer.
“If you are a consultancy you can generate more revenue by starting from scratch. So this idea is not going to be very popular with consulting organisations because it eats into their revenue, but from a vendor or IoT user perspective, it is good, because it reduces the time, the cost and the risk of a project.”
Industrial Internet Consortium’s reference architecture
Klement intends to take as his starting point the IoT reference architecture recently produced by the Industrial Internet Consortium, but said the aim was not to replicate it in any way. “We want to take specific use cases and build an architecture based on those that is independent of any specific technology.”
He said he would be relying heavily on his relationship with the IIC to realise the plan. “The IIC has 160 members and a test bed, if we want to do a reference implementation and we need the technical infrastructure to put the software on.” The IIC CTO, Stephen Mellor — who is also an adjunct professor at the Australian National University — due in Australia in October.
The IIC released its reference architecture in June saying it “provides a common language for the elements of industrial Internet systems and the relationships between them,” would “help developers decide which elements they need for their systems, enabling faster delivery of implementations” and “outlines key characteristics of industrial internet systems, various viewpoints that must be considered before deploying an Industrial Internet solution, and an analysis of key concerns for the Industrial Internet including security and privacy, interoperability, and connectivity.”
Dr Shi-Wan Lin, co-chair of the IIC’s Technology Working Group and principal engineer with the Strategy and Technology Office in Intel’s IoT group, said: “The Industrial Internet Reference Architecture has broad applicability across industrial and other IoT systems to drive that scalability and interoperability within the industry.”
Shortly after releasing its reference architecture, the IIC announced a collaboration with the IEEE’s P2413 Architectural Framework for the Internet of Things (IoT) Working Group to develop a “comprehensive architecture for an interoperable Internet of Things (IoT) around the world.”
Cisco pushing its own IoT architecture
Also pushing an overarching IoT reference model is Cisco, which launched the IoT World Forum Reference Model in 2014. According to Maciej Kranz, vice president of Cisco’s Corporate Technology Group, it organises all the components of IoT into layers, provides a graphical representation of IoT and all that it entails, and enables the creation of an ‘Open IoT’ system, with guaranteed interoperability.
Next step for the IoE Reference Architecture community project is to put in place the infrastructure to support collaboration among participants in the development of the architecture. He said he had approached a number of vendors of architecture modelling tools in search of a platform on which the architecture could be defined.
The only value of a common architecture is that it make it possible to effectively engage with a minimum of confusion. For example. If I talk about too many stands then this could be represented as IoT needs no more standards. This is not true. We probably don’t need more standards for connecting devices to networks, with more than 130 already. BUT we need to do some work in standardizing the way information is shared especially across different industries. A simple common architecture will help facilitate this sort of discussion. Developing yet another architecture is not particularly helpful. There are many and they already help to confuse the sectors most in need of help. Please don’t let us start debating whose architecture is the “best” as the result will be simply slower adoption from an even more confused market place.