Cactus Analytics

ANALYTICS. ANOTHER WAY TO INTERPRET INFORMATION.

The industry is shifting from traditional mileage-based, preventative maintenance regimes, and pivoting towards data-driven, condition-based, predictive maintenance that allows infrastructure companies to reduce the life-cycle costs of critical assets and components on the track.

Every infrastructure owner can benefit from Cactus Analytics for measuring operations with the ability to define key indicators for supervision, condition-based asset management and continuous improvement.

 JUST ANOTHER WAY TO LOOK AT THE INFORMATION

In applying the Cactus CCS integration platform and the Analytics toolset, a lot of new insights are created without any need for installation of additional sensors in the infrastructure

ASSET MANAGEMENT AND PREDICTIVE MAINTENANCE

Good asset management practice is crucial to mitigating potential costs in overrun penalties, especially as shifting requirements can derail even the most effective planning.

Cactus Analytics include capabilities such as:

  • Clustering (group data to search for similarities).

  • Correlation. Compare anything with anything.

  • Outlier and trend detection.

CONCLUDE ON GAIN

In applying the Cactus CCS and Analytics platforms it is possible to:

  • Measure performance before change.

  • Implement changes.

  • Measure performance after changes.

All to conclude on gain in the infrastructure

 

TRACK CIRCUITS

  • Normal distribution occupancy time with deviation analysis

  • Overspeed

  • Transients

AND GOOD MEANS TO VISUALIZE IT UNDERSTANDABLY

Through our Analytics framework, component groups (points, level crossings, etc.) can be visualized as over time, with each other, and correlated with other data from the Cactus CCS platform, in order to answer questions such as:

  • What components have the best operational data, with regards to purchase cost and amount of maintenance?

  • How does weather and traffic intensity correlate with the functionality of the track-side objects?

This graph shows a point with undetected, abnormal switch time, 15 seconds during 30 days, this entailed a “run-to-failure” approach that created a disturbance of traffic and a higher cost for getting back in operation. Detection of switch time is done by using already existing “dark data” from the traffic management system.