Who this is for: Reliability engineers, quality managers, inspection supervisors, IT architects, and anyone tasked with automating visual inspection across physical assets. Whether you are evaluating MVI or already have cameras in the field, this series is your implementation playbook.

Your Inspectors Are Missing Things

Here is a number that should keep you up at night: human visual inspectors miss 20-30% of defects on average. Not because they are bad at their jobs. Because the human eye fatigues after 20 minutes of sustained inspection. Because lighting changes. Because the hairline crack at asset number 847 looks a lot like the shadow at asset number 846.

Now here is the number that should get you out of bed: organizations deploying Maximo Visual Inspection report 81% inspection time reduction and defect detection rates above 96%.

The gap between those two numbers is not science fiction. It is computer vision, trained on your defect types, deployed to your cameras and drones, connected to your work orders.

This series tells you exactly how to close that gap.

What This Series Covers

Twelve blogs. Parts 1-10 build on each other. Parts 11 and 12 are companion references you will reach for repeatedly.

  SERIES ROADMAP
  ==============

  CORE SERIES:
  Part 1:  Introduction .............. What MVI is and why it matters
  Part 2:  CV Fundamentals ........... How computer vision works for assets
  Part 3:  Deployment & Infra ........ Deployment options, GPU, licensing
  Part 4:  Installation & First ...... Prerequisites, setup, first project
  Part 5:  Building First Model ...... Data collection to trained model
  Part 6:  Deploying Production ...... Inference pipelines and thresholds
  Part 7:  MVI Mobile (iOS) .......... Core ML, on-device inspection
  Part 8:  Edge, Drones & Field ...... MVI Edge, MQTT, drone integration
  Part 9:  Integration with MAS ...... Manage, Health, Monitor, Predict
  Part 10: Best Practices ............ Governance, retraining, scaling

  COMPANION REFERENCES:
  Part 11: REST API Reference ........ Complete endpoint docs and automation
  Part 12: Troubleshooting & FAQ ..... Every problem diagnosed and solved

The Series at a Glance

Part 1: Introduction to Maximo Visual Inspection

Tier: Free | Read Part 1

What MVI is, where it fits in the MAS ecosystem, and why computer vision is transforming asset inspection. Industry use cases across manufacturing, utilities, oil and gas, and transportation. The business case with real numbers. No hype, just what it does and what it does not do.

Part 2: Computer Vision Fundamentals for Asset Managers

Tier: Developer | Read Part 2

How computer vision actually works, explained for people who manage assets, not algorithms. Image classification, object detection, and segmentation demystified. Transfer learning and why you do not need millions of images. The no-code model building promise and where it holds up.

Part 3: MVI Deployment & Infrastructure

Tier: Developer | Read Part 3

Five deployment options (SaaS on AWS, SaaS on IBM Cloud, on-premises OpenShift, Azure, or client-managed RHOCP). NVIDIA GPU requirements and sizing guidelines. Storage configuration. AppPoint licensing at the Premium tier. The decision matrix for choosing your path.

Part 4: Installation & Your First MVI Project

Tier: Developer | Read Part 4

Prerequisites checklist (infrastructure and software). MAS 9.0 and 9.1 changes that affect your setup. The installation walkthrough from operator deployment through GPU configuration. The 30-minute quick start project to verify your end-to-end pipeline works.

Part 5: Building Your First Inspection Model

Tier: Developer | Read Part 5

The complete workflow from data collection through trained model. Image requirements, labeling strategies, training configuration, and accuracy metrics. How to interpret precision, recall, and mAP. When to iterate and when to ship.

Part 6: Deploying Models to Production

Tier: Developer | Read Part 6

Getting models out of the lab and into the field. Inference pipelines, confidence threshold tuning, handling edge cases and false positives. A/B testing models, rollback strategies, and production monitoring. The deployment checklist nobody gives you.

Part 7: MVI Mobile: AI-Powered Inspection on iOS

Tier: Developer | Read Part 7

MVI Mobile for iOS/iPadOS using Apple Neural Engine and Core ML. Which model types export to mobile (only three). On-device inference for offline field inspection. Device requirements, workflow, and mobile best practices. Why there is no Android version and what to do about it.

Part 8: MVI Edge, Drones & Field Deployment

Tier: Developer | Read Part 8

MVI Edge for real-time inference at the camera location. MQTT alert pipelines to Monitor. Edge diagnostics dashboards. Hardware selection from Jetson Nano to GPU servers. Drone integration patterns. Field deployment strategies for connected, intermittent, and air-gapped environments.

Part 9: Integration with MAS Applications

Tier: Developer | Read Part 9

Connecting inspection results to Maximo Manage for automated work orders. Feeding visual condition data into Health for scoring. Triggering Monitor alerts from visual anomalies. Enriching Predict models with inspection imagery. The closed-loop inspection architecture.

Part 10: Best Practices, Governance, and Scaling

Tier: Developer | Read Part 10

Model lifecycle management, retraining schedules, and drift detection. Quality governance frameworks. Scaling from one asset class to enterprise-wide visual inspection. The organizational change that makes or breaks AI adoption.

Part 11: REST API Reference

Tier: Developer | Read Part 11

The complete MVI REST API reference. X-Auth-Token authentication, every endpoint documented with examples, query parameters, data augmentation API, automation patterns for CI/CD pipelines, and the IBM vision-tools CLI. The developer guide for programmatic visual inspection.

Part 12: Troubleshooting & FAQ

Tier: Developer | Read Part 12

Every common MVI problem diagnosed and solved. GPU configuration failures, training issues, deployment errors, mobile limitations, edge connectivity, integration debugging, and MAS 9.0/9.1 migration gotchas. Plus 20+ frequently asked questions with verified answers from IBM documentation.

How to Use This Series

If you are evaluating MVI: Start with Part 1 (free) and skim Part 3 for deployment options. Then Part 5 to understand the model-building effort.

If you are starting implementation: Read Parts 1 through 5 in order. That takes you from zero to a trained model. Part 12 is your troubleshooting companion.

If you have models in production: Parts 6 and 9 are your operational guides. Part 10 covers governance and scaling.

If you are building automation: Part 11 is your API reference. Use it alongside Part 9 for integration pipelines.

If you are planning mobile or edge: Part 7 covers iOS/iPadOS mobile deployment. Part 8 covers edge devices, drones, and field patterns.

If you are struggling with model accuracy: Part 5 covers training best practices. Part 6 covers threshold tuning. Part 10 covers retraining pipelines.

If something is broken: Part 12 has every common problem diagnosed and solved, from GPU errors to integration failures.

The 8 Commandments of This Series

  1. Domain expertise before data science. The person who knows what a cracked insulator looks like builds better models than the person who knows gradient descent.
  2. One defect type first. Not twelve.
  3. Labeling quality determines model quality. Every time.
  4. A detection without a work order is a wasted inference.
  5. Models drift. Your retraining pipeline is not optional.
  6. Edge deployment changes everything. Plan for disconnected reality.
  7. Good enough at 93% accuracy in production beats perfect at 99% in the lab.
  8. Cameras are cheap. Missed defects are expensive.

Start with Part 1. Understand the landscape. Then go train something that catches what your eyes miss.

This series is part of TheMaximoGuys technical content library. Each part is written by practitioners who have deployed Maximo Visual Inspection in production environments across manufacturing, utilities, and energy. Parts 11 and 12 serve as companion references you can bookmark and return to throughout your MVI journey.

TheMaximoGuys | Enterprise Maximo. No fluff. Just results.