WBR Field Service Medical Conference Roundtable, March 8, Berlin
by Oliver Lemanski, General Manager, EMEA

Today most machines are connected to a network and are capable of recording their health and status in logs through machine data. While many companies are starting to use machine data more and more to make business decisions, there are challenges that need to be overcome in order to fully benefit from it. Machine data acts as the “Voice of Product” and when combined with customer, service and repair data may be able to provide insights that can be used for reduction in operating costs, revenue generation and better customer experience.

On March 8th 2016 OnProcess Technology hosted two roundtable discussions at the WBR Field Service Medical conference in Berlin. The purpose of the roundtables, which followed on from similar sessions hosted by OnProcess Technology at the sister Field Service Medical conference in the US in February, was to understand how the Medical Technology service industry is harnessing the potential of machine data (& analytics) and to discuss the challenges and opportunities that face manufacturers over the coming 3-5 years.

Usage: Who’s listening to the machines and what do they do with the data?

Of the 25-30 Medical Technology companies that participated in the discussion, approximately 80% were actively using machine data routinely in their Field Service and supporting operations. machine data in med tech companies field serviceThe bulk of these were using this information as a data feed or input for field engineers to better diagnose the problem and act accordingly.  This diagnosis is either conducted remotely or on site when engineer accesses the data logs.

Less than 15% of the participants mentioned that they routinely applied predictive analytics to the data. Those that did use predictive analytics to identify potential service issues and schedule a pro-active maintenance event or remote fix. This included the use of decision trees to speed up time to solution.

None of the participants used analytics on machine data to predict or assess the impact to other supporting functions of the service supply chain, such as spare part usage, or returns.

Challenges: What’s holding manufacturers back from doing more with the data?

This topic raised an interesting array of issues faced.

Challenges machine dadta and analytics

  • Data regulations: Whilst the focus of most data compliance regulatory guidelines is the patient data which isn’t necessarily needed for the purposes of device service, manufacturers voiced various issues related to this topic:
    1. Separation of data: the majority of the participants voiced that they could not separate the machine performance data from patient data.
    2. Access to data. In response to the security threat and tight data guidelines, several manufacturers face significant challenge with hospitals in being able to connect with their devices remotely. There was a high degree of variation on this topic, with some manufacturers not feeling that this was an issue.
    3. A few manufacturers felt that even if they could routinely access their machine and separate it from patient data, the degree of risk from data hacking or penalties from regulatory bodies was too high.
  • Company politics/silo mentality: In most cases, the field engineers are managed by local countries or markets within a manufacturer. Many manufacturers voiced that this creates a challenge in being able to centrally receive and process machine data and then create local action. The same was voiced in other functions such as supply chain for parts dispatch or R&D for future product design.
  • Technology: Several manufacturers felt they had either insufficient hardware, IT integration or processing tools to effectively manage machine data. This included data feeds not being stable or reliable. The sheer volume of data was also mentioned as technology challenge, with one manufacturer stating that they are struggling to conduct analytics on a 12 TeraByte (and growing) data pool.
  • FSE engagement: There was a noted impact to Field Engineers of improved use of machine data, remote diagnostics, and scheduled events from predictive analytics models. One manufacturer saw this as de-skilling the field engineers, whilst another saw it just providing the engineer with more information to work with. However, the majority agreed that younger engineers were more ready to adopt to the new ways of working, whilst elder engineers found it more difficult. One manufacturer explained that they had changed their recruitment policy accordingly.
  • Product design: The product design cycle for medical technology is long, meaning even the newer devices have already out-of-data data communications design.
  • Data not used: Two manufacturers mentioned that whilst they were able to collect more data from machines, they did not see an uptake of usage of this data.

Although it was not mentioned as a challenge by the participants, an additional issue may be focus; less than 10% of the field service professionals asked had improving data usage or analytics on their top 3 personal objectives.

Opportunities: What opportunities are there in the coming 3-5 years?
Opportunities machine data use

  • R&D: Many of the participants saw the machine data and analytics as an opportunity to improve future product and part design.
  • Improve service proposition:
    • Improve service experience: make more friendly and more tailored to each customer
    • Move to outcome-based model, where the manufacturer can move from tradition “uptime” monitoring to metrics more in tune with a hospital’s goals. There was open debate as to what a relevant outcome metric could be.
  • Do more with the data:
    • Increase % of service events through predictive analytics on machine data
    • Use data analytics to improve the supply chain managing spare parts and inventory planning
    • Potential use in quality surveillance for regulatory bodies
  • Improved remote support: With improved machine data visibility and analytics, it is a logical conclusion to suggest an opportunity to improve remote support (both hardware and software) capabilities.
  • Diagnosis moving to cloud: Two manufacturers, both of whom were already using predictive analytics to interpret machine data, described a market opportunity of cloud based diagnosis and action technology, which may or may not be in-house or outsourced. This again highlighted the issue of the data integrity and regulatory issues, which would prevent data leaving the EU, suggesting that the appropriate cloud technology would have to remain EU based.

Thought leaders: Who’s leading the field and what can we learn?

Of all the participants at the round table, one stated that they currently scheduled 10% of their service events through predicted failures interpreted from machine data. They also stated that in the coming 2-3 years that number could rise to as much as 25% of total service volume, whilst the majority of other manufacturers suggested a lower proportion for their business. The early adopter also stated that they saw this topic as a key growth enabler for their service business, but that in isolation it was not a “silver bullet” solution for the challenges they face in the field.

Conclusion

While many medical technology companies collect machine data, there are clear challenges that need to be overcome to be able to successfully analyse the data and take meaningful actions on predicted outcomes. Relatively low volumes (and high value) of installed products and the delicate nature of the relationship between manufacturers and their customers further slow the change process. However, early adopters are already seeing the benefits of this approach and the next 3-5 years could see a rapid evolution of how the machine data is being managed.

 

Thanks: OnProcess Technology would like to thank all the participants of the roundtables for their input and discussion.