The Device Chronicle interviews Alexander Kakavas, an expert in IVD diagnostics machines and IoT transformation.
Alexander has a strong expertise and background in medical diagnostics, medical technology and IoT. Alexander has experience in the international division of Roche Diagnostics in running transformational IoT projects such as how embedded gateways enhance the software management and maintenance of medical devices.
He begins the interview with his thoughts on how software is managed on large medical devices. He says “For much of the software deployment, the process is manual, engineers on site start the installation of the software on the machines depending on hardware and requirements of customers. The fleets are large – in the range of hundreds of thousands – that are connected to the Internet via a gateway in most cases. In exceptional situations such as in the defense sectors, Alexander explains, that medical test equipment would be completely isolated from Internet connectivity. In these scenarios, “there is no access, everything is offline and manual.”
Normally, every machine in the field is connected by gateways that are based on embedded Linux, that gets updated itself or manually “as frequently as possible and necessary.” There is also a need for maintenance. The gateway can be embedded on the device or connected by a serial connection. All the machines need software updating. They are developed mostly by manufacturing, mechanical engineering partners while the medical content for the machine comes from the medical company.
A major challenge for automated firmware updates is that there are many legacy machine models out in the field. This means that it is not easy to embed new digital technologies, or have seamless connectivity, and to have the capacity to perform proactive or predictive maintenance based e.g. on AI, where you need the input of the machines. Alexander explains “Many machines have been in the field for several years but are still robust. They perform their intended function well and you cannot expect these machines to have the electronics and digital capabilities to provide a lot of inputs for advanced analytics.”
Alexander explains that the OTA software updates strategy is often dependent on the portfolio of the medtech company in question. “Licensing and acquisition leads to a decentralization strategy and this creates a heterogeneous environment for the technology. In saying this, the tendency across the medtech industry is towards centralization to achieve homogeneity and economies of scale.”
The regulatory environment stays the same with the same levels of demand post-Covid. Countries such as China apply different regulatory rules which is costly for the medtech providers who must implement different rules and technologies. In the US and Europe, it is the same as before, there is caution around cybersecurity, the more exposure of the device, the more you need to protect them. There is normally hardware protection in front of the Linux gateway. “The secure Linux device is secured even further with another ring of defense.” There are certain machines where the machines are still not exposed. Alexander asks “How can these machines be updated if there is no connectivity interaction? Who will be in the machine room to perform the update? It lacks transparency and creates a lot of risk.” He says “In the US, the most advanced setups will be found, intrusion leads to cost and there is a realization of this.”
There are other barriers to adoption of OTA software updates: how advanced is the healthcare system? How is the market structured? For instance, Alexander explains that in China, there is an indirect channel with a continuous back and forth of devices from hospitals from distributors. This means that using a certain location for a machine is difficult, there is a need for a dynamic approach. Also, tracking firmware in machines is hard, or the hardware on the machine could be limiting the updatability of the machine. Alexander explains that there is a hope in the medtech industry that 5G will help with device connectivity. Often the machines are in locations that make it difficult to provide the services and LTE may not be available.
For the largest medtech companies, the cost of a physical visit to a hospital site is huge. Alexander estimates that it is a few hundred dollars per visit covering all costs associated with the engineer. So providing the update service remotely is what every company wants to achieve, and keeping the machines up and running 24/7 is another key benefit, and doing remote analysis so that an engineer is only dispatched when really needed.”
We wish Alexander well in his consulting work and new projects.