LAKSAHA | Ideas to Impact

OFFSHORE OPERATIONS / MATERIALS / MAINTENANCE

Working synthetic prototype · 25 October 2026

A controlled mobile materials workflow for offshore operations

Connecting equipment identity, drawing items, spare parts and work orders - with visible checks from issue through replenishment.

By Vijayaprakash Govindarajan · LAKSAHA Consultancy Services

Download two-page case study (PDF)Discuss a role or supervised pilot

The prototype uses deterministic controls and simulated roles. Live AI part identification and AMOS integration are proposed next steps.

Explore the newer Spares Readiness demonstration →

The operational problem

A technician needs a replacement part, but the manufacturer number may first need to be identified from a drawing. Equipment applicability, internal stock coding, bin location and work-order details must then agree. This demonstration makes those checks visible within one mobile workflow.

The verified owner test

5successful transaction events
1blocked approval attempt recorded
10usable seals after inspected storage

Owner-supplied phone screenshots and the exported JSON corroborate one synthetic run. QR inputs were demonstrated on the phone; scans are not separate events in the JSON.

StageObserved result
Identify and verifyEquipment EQ-DEMO-001; drawing SYN-DWG-001 revision A; mechanical seal SYN-SEAL-200 mapped to STOCK-002 and BIN-B02.
IssueFive seals issued against WO-DEMO-001. Usable stock falls from eight to three.
Request and approveRequest seven seals. Storekeeper approval is blocked and recorded; APPROVER-DEMO then approves.
ReceiveSeven received against DEL-DEMO-002 into quarantine, without increasing usable stock.
Inspect and storeInspection and correct-bin put-away increase usable stock to ten. Request #1 closes and receipt #1 is stored.

Practitioner contribution

Vijayaprakash contributed offshore materials and maintenance workflow knowledge, requirements, control choices and hands-on phone validation. The implementation was developed with AI assistance. This demonstrates the ability to translate operational requirements into a testable digital workflow.

What this demonstrates - and its limits

The working module connects drawing-item selection, part and bin verification, work-order issue, replenishment approval, quarantine and inspected storage. The exported record contains five successful transactions and one rejected approval attempt. Other stock balances remain unchanged.

It is a synthetic portfolio demonstration, not a customer implementation or a production-ready rig system. Roles are selectable simulations, not authenticated users. The local SQLite log is not tamper-proof. No live AI identification or AMOS integration is present. Issuing a part does not confirm installation.

Time savings, error reduction, downtime savings and customer demand have not been measured.

The next AI capability

Retrieve candidate parts from a permitted manual or drawing with page and item references, check model, serial applicability and revision, and require human verification before entering the controlled materials workflow. Approved substitutions must not be inferred from visual similarity.

How a supervised pilot would be measured

Any pilot needs agreed scope, authorised data, operator-approved equipment and integration, and named decision makers.

Relevant roles and services

Materials and maintenance systems, CMMS workflow improvement, asset and catalogue data quality, implementation requirements, user acceptance and training. A bounded engagement could assess one materials workflow and evaluate a supervised prototype against agreed acceptance criteria.

Contact VijayaprakashExplore the AADP-Ops repository

Repository link refers to the wider project. This page does not assert that the mobile extension has been merged there.

Demonstration recording script (90-120 seconds)
  1. 0-15 seconds: Explain the offshore replacement-part problem and state that this uses synthetic data.
  2. 15-40 seconds: Show equipment identity, drawing revision, seal item, manufacturer number, stock code and matching QR inputs.
  3. 40-60 seconds: Show the five-part issue against the work order and stock dropping from eight to three.
  4. 60-90 seconds: Show the recorded blocked approval, the separate approver, and receipt into quarantine.
  5. 90-110 seconds: Show inspected put-away, stock ten, closed request and stored receipt.
  6. 110-120 seconds: Explain the proposed source-cited AI lookup and invite a role or pilot discussion.

Use retained screenshots for the completed cycle, or a separate fresh synthetic database for a new recording. Do not reset or overwrite the evidence database merely to record a video. No video recording is included yet.