KEKO Experimentation Platform – Learning from end users and developers
KEKO experimentation platform is the first step towards a fully functional KEKO platform. It enables working with data coming from multiple sources, rapid development and integration of new technologies, and engagement with 3rd party developers.
KEKO is a dynamic ecosystem on a mission to make buildings safer, more fluent and more productive. Our way of working is based around experiments that we do in iterative and agile way. Nobody knows what the future will look like, but we humbly embrace the unknown. We want to test our ideas and learn quickly what works and what doesn’t. We are not only testing the technical feasibility, but also business viability and end user desirability.
But how do we experiment when we don’t have a platform available, or even a clear plan for developing one? Of course, we don’t wait for the platform to be finalized. We want to get our hands dirty now. We want to show quick prototypes and mockups to the stakeholders to understand where the value is. We want to show early ideas the developers to learn what kind of features, APIs and data they need to have in the platform to deliver the end user experiences. We want to understand if the platform should run in the cloud or in the edge, stand-alone or together with other platforms.
For this we need to have an experimentation environment. Something that we can build very quickly, using existing technologies, to get us started. Environment that then evolves during the project, enables our learning and feeds requirements to the longer-term development. We engage with customers and our developer partners simultaneously, and early on, to create a positive network effect between the development phase ecosystem, i.e. pilot customers and pilot developers.
Three Key Principles of Development Phase Platform
The KEKO team has identified key success factors to drive the innovation from technical development perspective for managing the pathway from ideas to a concrete digital building platform:
Enable rapid innovation of platform and APIs – in many case cases, APIs are perceived as the end product, creating a functional interface that enables systems to perform certain tasks or functions and that work through the agreed and well defined APIs. On the other hand, during the development the understanding of end users and developers’ needs are in constant change, it is essential to enable rapid innovation of APIs. Creating new “transient APIs” should happen in a matter of minutes, rather than days or weeks to make sure the innovation is not hindered by slow evolution of the APIs.
Enabling innovation based on full historical data vs functional APIs – The functional APIs are usually great for doing a specific function, but we have recognized a great need for giving full data access through controlled APIs for developers. Many new innovations come from artificial intelligence or machine learning, where the access to full data is important. The services developed on top of full data APIs can be migrated to functional APIs. For example, to create a predictive maintenance functional API you need access to full historical data to train an AI algorithm that can predict the current need for service, which again can be exposed as a functional API. Combining the full data API approach with rapid innovation approach, and engaging with 3rd party developers to rapidly create new innovative ways of using the data, is a key to identifying and creating new value adding APIs that can be productized for broader use.
Managing the development funnel – In order to manage the transient APIs – i.e. APIs you allow the be changed flexibly to maximize the speed of innovation – and production APIs – i.e. APIs where you optimize for stability and availability – you need to manage your development funnel properly. The key is to understand and distinguish a) areas where the customer needs and ideas are already well known and the APIs and the applications created using the APIs can be stabilized into commercial offerings and b) areas where the customer understanding and value creation is still evolving rapidly.
Building KEKO Experimentation Platform
KEKO experimentation platform is built using WhereOS. It enables quick integration of data from multiple sources and multiple formats. The data from digital buildings comes from multiple sources: sensors, building management systems, BIM models etc. All of them have different formats, protocols and ways of integration, depending on the vendor of the underlying solutions.
The experimentation platform enables rapid data processing and development of APIs using data, and to expose the source data from various systems in a consistent manner for developers. On top of the APIs the experimentation platform enables development of SDK features that add further value on top of the APIs, to create actual UI applications for end users.
WhereOS ecosystem is used as a service model that enables KEKO data and APIs to be published in the private marketplace area that is created specifically for the KEKO project. This enables KEKO partners and developers, for example, to take raw data from buildings and develop smart algorithms on top of it and release the intelligent data and APIs for other teams to use. This creates a positive network effect within the development ecosystem, as compared to “point solution” work, where each developer is working on their own solution, and the solutions do not build on top of each other.
Janne Öfversten, Senior Technology Manager / KONE Corporation
Jussi-Pekka Partanen, CEO / WhereOS