Jan van Zoest’s Post

Jan, thanks for sharing!

Defining a reference architecture and building the foundation is indeed essential for a company. But somehow there is never any time available to first create the platform as feature teams need to deliver :) So how to tackle the challenge: build a new foundation, keep the feature teams aligned, without blocking them and without creating too much technical dept... Any thought on this transition Jan?

Beyond supportive to faster first development, the utilization of reference architectures also caters to easier and more efficient maintenance of delivered applications and systems. Lack of willingness to comply to this principle is a bad practice in any enterprise application delivery, not limited to the world of medicine alone. I like your building metaphore, this extends also phases in which one needs to do (housing) reconstruction and extension: for best and thrustworthy results, the architects and developers must trust that common sense building principles are applied in the initial build, and that the (housing) foundation + plumping are delivered as you would expect on other constructions, and thus not too much influenced by the desire of the initial (system) architect to leave a personal trademark.

A thought-provoking article! However, in my experience, Enterprise architectures and Centralized data models find it difficult to cope with today's diverse applications. For instance, a reporting application needs a view of the data, which is often totally different from an ML algo, different from the needs of a transactional reconciliation algo. What's a central model going to offer? A standard that's different from all of them? Today's successful systems often rely on bounded contexts that are easily replaceable, not a set of long-term low-change rules. Would be interesting to see how things adapt in the next few years!

Architecture is only as strong as one's ability to communicate it. Strong article Jan, and congrats to the team on getting where you are today.

Cant emphasize & agree more on the need of "HealthSuite Reference Architecture" and the paradigm shift from technology focus to customer focus. Its analogous to a train journey where origin (current state) and destination(target state) are clearly envisioned, connected via guarded rails for a faster and safer commute. Its awesome to get recognized with "Best Digital Architecture in Healthcare" award.

Theo De Bruin

Cannot agree more! I find it very inspiring for my current challenges while defining an enterprise architecture within my own domain. I wonder whether the architecture framework for Reference Architecture of Philips Healthcare is Zachman or another new one defined and designed within Philips. I also wonder whether MBSE approach is used during the definition phase and projects can get benefit from models instead of a document based approach.

The only problem I see with a reference architecture is that it needs to be designed for a mature platform. Especially with mobile this is not the case, unfortunately. On mobile, developments on all levels are evolving so rapidly that it feels impossible to keep a reference architecture up to date. Both hardware and software are still immature. The whole eco system keeps changing dramatically, ranging from new hardware, security issues, privacy laws, build tools, programming languages and even ide's popping up every year or so. Off course, mobile is just a tiny part in the overal reference architecture, but it will always be time consuming to integrate.

Well written Jan. Cannot agree more having personally experienced the pitflalls of immature platforms impacting TTM. It is myth that guardrails curb innovation