I recently attended (virtually) the launch event of the AIoT user group, which was initiated by practitioners from Accenture, Bosch, Clariba, Deutsche Post, Evaco, mm1, Opitz, Recogizer, TH Köln and Tomorrow Labs. The event was hosted hosted by Ferdinand-Steinbeis-Institut (FSTI), a research institute for business transformation in the context of digitisation and networking. The FSTI is affiliated to Steinbeis University and aims to transfer knowledge from academia to business.
Starting with the basics: what’s AIoT? AIoT was characterised as the intersection of IoT (Internet of Things) and AI (Artificial Intelligence). These are both fast emerging technologies that will be very impactful in coming years. FSTI believes that the intersection of these two technologies is likely to both be particularly fruitful in terms of applications, and particularly challenging in terms of needing to blend together AI, communications, and hardware competencies.
The below diagram is a pretty good summary: the worlds of AI and IoT (cloud) typically work with ‘agile’ project methodologies, iterating fast and delivering continually improved software propositions. Conversely, the world of hardware assets tends to undertake developments using a ‘V’ model – gradually refining the specification for a hardware asset until it is deemed to be ‘final’, and then building it. These two worlds are not natural bedfellows, but AIoT forces them to cohabit.
In truth, this scenario is not exclusive to AIoT (it also manifests in plain old IoT environments), but the ‘A’ does add another level of complexity.
To aid both end users and practitioners in the emerging AIoT space the Ignite|AIoT project management framework has been published on the AIoT website: http://www.ignite-aiot.org/. This is a comprehensive framework that covers the topics highlighted in the graphic above, including reference materials related to each of the areas. It’s a comprehensive and well thought-through project management framework, and at the launch event it was suggested that it would take a user about 2.5 days to be ‘trained’ on the content. The overall aim is for Ignite|AIoT to become an established framework for reducing risks and accelerating time to market in converged AI and IoT projects. For background information, the Ignite|AIoT project management framework builds on the Ignite|IoT project framework, of which I was a co-author some five years ago.
Back to the overall user group (http://www.aiot-ug.org/), and the focus will be on ‘Smart Connected Products’, and general sharing of experiences and good practices through talks, case studies, boot camps, and of course the Ignite|AIoT framework which is intended to be a ‘living’ thing. The overall approach is summarised in the graphic below.
The launch event also included a presentation of ‘How AIoT is going to win the next world soccer championship’, delivered by Clariba. Well, this was a launch event held in Germany. Having said that, the beneficiaries of the AIoT project discussed were the Italian Football league, so this headline message will more than likely need to be re-spun when the eventual winners of the next world soccer championship emerge (be they German, Italian, or anyone else for that matter).
The actual football case study was quite interesting, and sophisticated. Apparently the currently deployed solution is the result of approaching 40,000 development man hours. It ingests diverse data sources ranging from real time match information to player medical history and even how players ‘feel’ on a particular day. Analyses support personalised training regimes, benchmarking of players, and the expected status of injuries and risks of further injuries for specific players.
Meanwhile the challenges highlighted by the project team will be familiar to many: integration between legacy systems, explosions of data, lack of maturity, communications between different departments, and so on. And fundamentally, it seems that football associations just aren’t focussed on technology!