At KONUX, we strive for innovation with the mission to make rail the mobility choice of tomorrow. To achieve that, it is crucial that we constantly develop and make our system better to accommodate various challenges that our customers might have. The base of such continuous innovation is the SaaS (Software-as-a-Service) solution model on which the KONUX system is based. With the KONUX SaaS solution, you can benefit from the continuously improving system and ongoing innovation.
Since our first launch of the trackbed health feature over 3 years ago, we have been gathering abundant feedback on how our customers are using the features and working to bring our system to the next level in terms of usability, actionability, and depth of insights.
We have improved and expanded the trackbed health features to enable you with even more accurate and actionable insights into trackbed health. In this Summer ’21 Release, we proudly present the trio of our trackbed health features – Trackbed Health State, Trackbed Prediction, and Trackbed Trend.
For trackbed health insights to be meaningful and useful, simply knowing the current condition is not enough. The system should allow a comprehensive understanding of how the condition is now, how it has been progressing in the past, and how it will progress in the near future. When such insights are combined, it finally enables you to detect the early signs of degradation, plan ahead for an inspection, and prioritize your maintenance actions.
The Trackbed Health State is the ultimate tool to guide your actions, showing you at a glance which assets need action right now (red state), close attention (yellow state), or no action at this point (green state). The improved feature now incorporates the predicted conditions of the trackbed in the health state so you can base your decisions on when it might cross the threshold, rather than how close it is now to the threshold.
This is to complement the current method of using a static threshold to base your decisions (e.g., ‘I should pay attention to switch A because the displacement is close to the predefined threshold at 3 mm’). It now allows you to plan your actions on the degrading assets based on the probabilities of them crossing the threshold in the near future (e.g., ‘there is at least 50% chance that the displacement of switch A will cross the 3mm threshold in the next 90 days, so I better pay attention to it’).
The Trackbed Prediction feature allows you to estimate the future state of the trackbed so you can weigh the urgency of which ones need to be prioritized for actions. This is not to foretell how exactly the trackbed will look on a specific date in the future, but rather to give you a realistic indication of the range of displacement the trackbed will be at in the specified timeframe, i.e., 90 days.
This feature now provides much more accurate and stable predictive insights on trackbed conditions based on the improved predictive model. The advanced data pre-processing excludes irrelevant events and noise, basing the prediction only on meaningful signals. Also, the model requires and uses far more data before it calculates the predicted values, making the insights statistically more solid. Our prediction model testing against the historical customer data reveals an accuracy of over 90%.
The new Trackbed Trend feature provides an additional way to assess the trackbed health based on its retrospective trend of the past 90 days. This can complement your decision-making as it helps you understand how the trackbed has been behaving in the past and take this into account to assess the current and predicted health conditions.
This feature helps you assess the full scope of the trackbed health, as knowing the current displacement level is often not enough due to differing degradation speeds. For example, the displacement of switch A can stay at 2.8mm for months, while the displacement of switch B can increase from 1 mm to 3 mm threshold in a matter of weeks. Also, the Trackbed Trend can provide complementary insights into how the condition might progress when, for example, the prediction may not be available due to insufficient data.
Data is there to guide your actions. The combination of the two improved features Trackbed Health State & Trackbed Prediction and a new feature Trackbed Trend provides you with a comprehensive and accurate understanding of the trackbed health conditions of the past, today, and the future. And this, together, enables more actionable insights than ever to guide your actions.
Would like to know more about this article? Get in touch here.