Skip to content

Materials and data

Public documents used in Lab 4

All documents are published by their owners; do not redistribute the files outside the training workspace. Copies are bundled in this repository at data/lab4-documents/ and copied into each participant's volume automatically by the lab_data setup notebook. The links below are where they came from, each checked and current, in case they're ever updated and need refreshing:

Document Owner Pages Link
FANR-REG-12, Regulation for Emergency Preparedness and Response for Nuclear Facilities FANR 20 PDF
FANR-REG-13, Regulation for the Safe Transport of Radioactive Material FANR 195 PDF
FANR-REG-17, Regulation for the Certification of Operating Personnel at Nuclear Facilities FANR 14 PDF
FANR-REG-30, Regulation on Technical Services Related to Radiation Safety FANR 8 PDF
IAEA SSG-30, Safety Classification of Structures, Systems and Components in Nuclear Power Plants IAEA 44 PDF
IAEA Nuclear Safety and Security Glossary, 2022 (Interim) Edition IAEA 248 PDF

FANR regulations carry a "Public" classification marking. IAEA publications are free to download and copyrighted; they are used here for training only. The glossary replaces the older "IAEA Safety Glossary" (2018 Edition), which its own cover page marks as superseded — use the link above, not the old one.

Skip anything marked "for official use only" (for example the regulatory guide for FANR-REG-08).

Limits: each file must be under 100 MB and 500 pages, in pdf, docx, pptx, txt or md.

Capstone memo templates, used in Lab 4 (Knowledge Assistant)

Four short, fictional memo templates and procedures — leave requests, change requests, procurement approval, and shift handover — written for this course, not copied from any real organisation. Bundled in this repository at data/capstone-memo-templates/ and copied into each participant's volume, in a memo_templates subfolder, by the same lab_data setup notebook that copies the Lab 4 regulation PDFs. Used by the Memo and approval helper capstone case (see Capstone).

Synthetic datasets

Generated by data/generate_synthetic_data.py in this repository. Nothing in them is real.

Table Description
work_orders 2,000 preventive and corrective work orders across four units, with status, priority, dates, asset and technician
assets 120 assets (pumps, valves, transformers, sensors) per unit and system
technicians 40 technicians with team, shift, and a masked phone number (see the data-masking exercise in Lab 2)
it_tickets 500 IT service desk tickets, about 20 percent in Arabic
incident_reports 60 short narrative reports for the extraction demo
meeting_transcripts 10 short standup-style transcripts with action items, used in Lab 3 and the Meeting assistant capstone case
monthly_kpis 12 months × 4 units of KPI figures (availability, planned maintenance completion, corrective work orders, safety incidents, budget variance), used in the KPI and annual operating report capstone case

Generated fresh by the lab_data notebook, which every participant imports straight from GitHub and runs themselves at the start of the training (see Before you start) — it creates the data directly in their own schema, nothing to download or upload by hand. The same notebook removes it again at the end (switch it to Clean up my data). See Trainer notes for the one-time setup this depends on.