Start with the maintenance meeting question

Import the HS asset-management scenario and separate the four evidence layers needed to prioritize maintenance.

Question for this chapter

What evidence do we need to prioritize maintenance for ten assets?

Why this matters now

Starting with seven input datasets and five algorithms makes it easy to complete the runs without understanding how each output informs a decision. First separate current condition, remaining time, relationship context, and execution constraints.

Try it

hs_asset_management.zip Download(97 KB)

Select Collections in the left sidebar. On the Collections page, select More (⋯) → Import in the upper-right corner and upload the ZIP. If validation reports no conflicts, select Start import and wait for completion.

Successful result

The collection tree shows hs_asset_management. Open equipment_master and confirm ten rows from EQ-001 through EQ-010. The other six input datasets are populated, while the five output datasets remain empty until you run their pipelines.

Korean Portal showing input datasets and analysis assets in the HS asset-management collection — The Portal capture is in Korean. Confirm the seven inputs and the analysis assets that later populate five outputs.

Read the result

The evidence needed for the first maintenance choice has four layers.

Evidence	Question it answers	Asset to inspect
Current condition	What do several checks say about the asset now?	`health_index_scores`
Remaining time	What range follows from failure history and operating conditions?	`rul_estimates`
Relationship context	Which subsystems, sensors, and inspections connect to the asset?	`hs_equipment` ontology
Execution constraints	When can work fit within budget and daily crew capacity?	`maintenance_schedule`

Deep dive — assets registered by the import

Asset type	Count	Contents
Collection	1	`hs_asset_management`
Datasets	7 inputs / 5 outputs	Equipment, sensors, inspections, maintenance, DGA, failures, operations / anomalies, HI, RUL, schedule, report
Code assets	6	Thermal anomaly, HI, RUL, schedule, ontology, report
Pipelines	5	Health, reliability, maintenance plan, ontology, periodic report
Ontology	6 entities / 6 relations	Equipment, subsystem, sensor, inspection, defect, maintenance and their relations
Knowledge / dashboard	1 each	Maintenance manual and equipment health overview

The input row counts are 10 equipment, 1,000 sensor readings, 33 inspections, 20 maintenance records, 25 DGA samples, 33 failure-history records, and 500 operating-condition rows. Pipelines populate the outputs in the following chapters.

Next decision

You separated the evidence layers. Next, combine several condition clues for EQ-005 into one HI and ask whether the lack of a thermal anomaly is enough to lower its priority.