Last month, I attended a two-day IoT-A workshop during IoT Week in Helsinki. The goal of the workshop was to showcase the various IoT research projects that are jointly funded by industry and the EU’s FP7 research program. The quality of the projects on display was amazing and I could not possibly do it justice in the space of a blog post. Still, here’s a partial list of what I saw:
- Horizontal open platform for IoT
- To learn the intent of a user requires a horizontal approach
- This horizontal approach leads to context awareness
- Part of Future Internet Public-Private Partnership (FI-PPP) program
- Composed of Virtual Objects, Composite Virtual Objects and Service Layer
- User characteristics + situation awareness = intent recognition
- Linked sensor middleware
- Data management instead of infrastructure management
- Uses information interoperability and linked data to enable automated composition
- Manufacturing automation
- Uses XACML and extends it for linked data
- Probabilistic registration of things
- Registration decisions are based on existing density and coverage requirements
To get a more complete picture, you can find all the presentations from the workshop here.
There were two key insights I took away from this workshop, both of which had to do with subtle similarities shared by all the projects.
First, sitting in and listening to the various presentations, I was struck by one particular similarity: at the core of each use case was the desire to make better-informed decisions. I’ve tried to capture what I call the core motivation of IoT in the picture below.
The identity of the user or thing combined with the temporal and/or spatial context based on real-world knowledge and data from the past can allow us to make better-informed decisions for the future. I think this holds for both the smart coffeemaker and the smart city.
My other insight had to do with the surprisingly similar characteristics of the various presented IoT applications. I tried to capture these characteristics in the picture below.
At the heart of the applications lies data – lots of data. But Big Data has two siblings: Fast Data and Open Data. The applications are graph-structured based on the relationship of things to each other and to me. They are event-driven rather than transactional and they are compositional.
What do you think? What kind of similarities do you see between the various applications?