Medical devices need to operate around the clock reliably. Medical devices are not only life-improving but mainly life-sustaining. So it is important to reduce unexpected outages and avoid very expensive emergency visits of field service engineers.
Even from an experience with a customer delivering devices for smaller clinics to measure blood values or urine values, I can say that it is becoming more and more important that wearables measuring blood pressure are reliable and do not fail so that there is no demanding timeframe needed for their maintenance. The same applies for more complex devices even at a higher degree.
Predictive Maintenance is based on data generated and transmitted from medical devices whereby manufacturers can monitor the status of a medical device and recognise if it is operating normally or abnormally and proactively predict a possible failure and plan to intervene before the device breaks down. Sensors may record the environmental conditions such as the temperature of the medical device, the humidity, any kind of vibrations, etc.
Experience from a recent project
I have been involved in one project for more than a year focused on building up a big data, AI and service analytics platform. The data scientists are looking at all the data and exploring new AI-powered predictive models to predict future failures on device components associated with big cost impacts.
Optimising maintenance schedules
If you want to predict future outages, you need data. That is one of the reasons why many companies invest in the IoT, which creates data from various sensors. Imagine a smartphone. How many sensors does one have? It is amazing; such a small device has approximately 6–8 different sensors. With these sensors, data companies are building new applications.
If you consider a medical device, there are so many factors that can impact the functioning of a product. Sensors measure humidity, temperature in the operating area and so on. These all influence the time between maintenance events. In some countries, some elements fail quicker than others, and this need to be predicted and mitigated. The trick with predictive maintenance is to predict as precisely as possible when a system break will hit your customer; it should not predict the future event too far in advance nor in the reactive phase of handling.
Improvement of Customer Satisfaction
Automated data collection and processing makes it possible for a service provider to extract insights from the devices’ statuses, operations and behaviour. These insights unlock a tremendous potential: Manufacturers of medical devices can grow by refreshing their current product-related offerings by adding new digital services. According to a BCG study, companies that can successfully reposition their service capabilities stand to generate substantial standalone revenues with profit margins in the 35% to 40% range—more than 50% for best-in-class performers—while also improving customer satisfaction and providing a source of competitive differentiation.
Main benefits of Predictive Maintenance
First-Time Fix Rate improvements
Higher customer satisfaction rates
Less travel costs and less capital locked in stocks
What does the customer need to bring along at the start?
- Quality and quantity of data
- Data engineering aspects need to be given
How ERNI can support in this area
- Business & Technology Consulting
- Innovation Services
- Project Management
- Architecture Advisory
- IoT Engineering, Big Data Engineering, ML Engineering
- UX, Front-End Developers Agile Coaching
