Forcing open interfaces? Two examples

The Federal Ministry of Health (BGM) has called for open interfaces in medicine - both for software and hardware, such as pacemakers.

This is not the first time that openness has been demanded by decree. Another example is the European Train Control System (ETCS), which enables the cross-border movement of high-speed trains.

The advantages and disadvantages, opportunities and risks are examined below.

Interfaces - why actually?

If we develop without explicitly defined interfaces, we end up with a monolithic solution. Interfaces, on the other hand, enable decoupling. This has many advantages if the interfaces have been defined cleanly and completely:

• Complexity becomes manageableas each system separated via interfaces can be developed and tested independently of the overall system.
• Technical dependencies are reducedas another system can simply be connected to a clearly defined interface.
• It is easier to react to changeas changes often do not have an impact across interfaces

But interfaces also have Disadvantages:

• The clean and complete definition of interfaces generates additional effort. This is only justified if the interface is also used as such (i.e. different systems use the interface).
• Incompletely defined interfaces can cause unexpected problems. Markus Bolli reported on such incompleteness in the ETCS standard. In his keynote speech in 2016, he reported that various manufacturers had interpreted the standard differently and were therefore not fully compatible.
• Interfaces have the following effects negative impact on stability and size out. Hence the trend from computer and smartphone manufacturers, soldering more and more components in their devices and gluing the devices together.

Industry only interested in open interfaces to a limited extent

Anyone who worked in IT in the 90s may remember the "war" between the manufacturers of inkjet printers and users who preferred to refill cartridges themselves instead of buying new ones for a multiple. In this example, the manufacturer is definitely interested in interfaces, but not open ones. The money was made with the cartridges.

However, from an economic point of view, an open interface can be interesting as long as the manufacturer controls the standard and dominates one side of the system. For example, the software interface to Android is open, but Google controls the app store (Play Store) through which the majority of apps are installed. Google uses a clever license structure to ensure that it is inconvenient to bypass the Play Store. These structures are even more pronounced at Apple.

And finally, there are also situations in which the industry is grateful for open standards. Namely, when it does not affect their own products, but the necessary infrastructure. For example, hardly any printer manufacturer would be interested in not supporting the A4 paper format.

Back to the railroad

With the introduction of high-speed trains in the 1980s, the issue of cross-border rail traffic became increasingly important. In the 1990s, the EU decided to promote the creation of a catalog of requirements for interoperability in high-speed transport. This was also positively received by the industry.

The project took off and has now been successfully established in Europe. However, two problems were barely noticed by the public:

• Until the 2010s, there were still Significant problems with interoperability. The problem could only be contained through concrete measures. In particular, it was so bad in the 2000s that some trains carried a separate ETCS control unit for each country in which they were allowed to run. Different versions of the standard further exacerbate this problem.
• The introduction of ETCS is associated with considerable costs. The lack of interoperability has added to the already high costs.

Nevertheless, the bottom line is that we can draw a positive balance here, because the result is an attractive product that still meets the safety requirements of the rail industry.

The fact that a common standard is also in the interests of train and track operators has certainly helped here: train operators can expand their services across borders, while network operators get more capacity utilization if more operators can use their network.

The standard is not quite as attractive for component manufacturers: Because complete interoperability would mean that customers (train and network operators) could buy components where the offers are most attractive. I don't want to comment on whether the incompatibilities observed in the early 2000s have anything to do with this or not...

Open interfaces in the healthcare sector?

In view of what we have learned in the past, what can we learn from the Call from the Federal Ministry of Health for open interfaces in medicine? The call is initially about open interfaces in software solutions, which suggests the exchange of patient data, prescriptions and the like.

But it will certainly not stop there. For example, the Greens have asked whether a Standard for parts implanted in the body makes sense. This already shows that we are dealing with an immensely larger scope than with ETCS.

Particularly in view of the advantages and disadvantages mentioned above, it could be difficult to get the industry to develop such a standard. For example, the frequency of connecting another component (pacemaker) to an existing interface (pacemaker electrode) is likely to be very low. In order to realize this independently of the manufacturer, the cost-benefit calculation is likely to be poor.

In practice, something like this could work if a strong player in the market creates a standard itself, demands it for its own products and builds a business model on it. Perhaps there is a manufacturer of electrodes that suggests a standard to the manufacturers of pacemakers to connect them to their own devices.

However, the situation is different for patient data or electronic prescriptions: Here it is to be expected that there are numerous producers and consumers of data. There is a completely different dynamic here.

But even there, it is all too easy to get bogged down. A government mandate certainly makes sense in order to initiate major transformations. But for this to work, the interests of all stakeholders must be taken into account. The discourse so far seems to be well on the way to getting bogged down in different directions. Let's hope that we can regain focus in order to improve the healthcare system in the long term.

Photo by Ani Kolleshi on Unsplash