Deep dive round table discussion – moderated by KONUX
Date: 7 November 2019
- Changing processes: How do you change the processes from pre-scheduled maintenance and inspection cycles to usage-based / predictive maintenance practices?
a. What are the barriers today?
b. Who needs to be involved in changing the process?
c. What data, procedures, do we need?
- From equipment to SaaS: How do you buy a SaaS solution today?
a. What are the barriers to buying a SaaS solution?
b. What needs to change to make it easier for railway companies to test and tender SaaS solutions?
- Data policy: What are the most important guiding principles when designing your data policy?
a. Do you already have a data policy in place when working with IoT/ monitoring companies, intelligence providers, etc.?
b. What is your purpose of data, core business, solve a specific problem, enable learning, etc?
c, How do you ensure collaboration between relevant stakeholders?
- Key assets: Which railway assets do /would you address first with Predictive Maintenance?
. Where are your blind spots?
i. In track?
ii. In the infrastructure as a whole?
- Quantifying success: How do you measure the success of a predictive maintenance project?
a. What is crucial when building a business case?
b. How do you ensure you also consider the benefits that are harder to quantify within the timeline of a pilot?
- Ground truth: how do you ensure “ground truth” to train machine learning algorithms.
a. What are the barriers to providing “ground truth” to suppliers?
b. What needs to change to accelerate learning and value creation?