AI inspections for IBM Maximo
RapidEye reads the images your technicians attach to Maximo work orders and inspection forms, and returns a condition finding tied to the asset and location they came from.
What a Maximo work order carriesOverview
A Maximo work order is a complete record of what was done, and a very incomplete record of what things look like. It holds the asset, the location, the job plan, the labour, the parts, the failure codes and a status that moves from waiting for approval through to closed. Images ride along with it, on the record itself and on inspection form questions, and once the status reaches COMP or CLOSE almost nobody opens them again. RapidEye reads those images, together with the asset, the location and the time, and returns one finding per issue it sees: damage, wear, a cleaning miss, a missing part, a safety problem, with how bad it looks and the images behind it.
How it works with IBM Maximo
Maximo already sorts work by the reason it exists, and the images from each kind of work answer a different question. The checks follow that division.
- Corrective work orders. A technician photographs the fault, fixes it, photographs the repair, and sets the record complete. RapidEye reads both and says whether the surrounding area is in the state it should be: the leak stopped but the ceiling tile is still stained, the pump is running but the bund is full.
- Preventive maintenance routes. The same asset comes round every quarter with the same job plan, so its images stack up in a line. Corrosion that grew over four visits reads as a trend rather than four unremarkable photos filed months apart.
- Inspection forms. Where a question asks for a file upload, the photo is attached at the question that raised it. RapidEye reads the picture next to the answer, so a Pass with a photo showing a cracked fitting comes back as a finding instead of a green tick.
- Locations and spaces, not only equipment. Plant rooms, corridors, roofs, washrooms and loading bays all sit in the location hierarchy. What the photo caught in the background of an asset shot is condition data about the space, and it is read as such.
- Contractor and vendor work. When the person closing the record is the person billing for it, the images are the only independent account of what was left behind.
Each finding names the asset, the location, what kind of problem it is, how serious it looks, the images it came from, and a suggested next step. It lands in the RapidEye dashboard, and in a Slack channel or an inbox for the reliability or facilities team if you want it there. Your planner decides whether it becomes a follow-up work order, a defect, or nothing. If your teams capture something other than still photos, video walkthroughs of a plant room, 3D scans of a space, or an inspection format of your own, that is a conversation with us rather than a limit: the reading is not tied to one kind of file, and the standard it reads against is yours.
How to connect IBM Maximo to RapidEye
- A Maximo Manage environment with your assets and locations in the hierarchy. Tell us which sites and asset classes are in scope. On a large estate this is usually one region or one asset class first, and the rest of the hierarchy stays out.
- Wherever the images already are. Attached to the work order, or attached to inspection form questions, or both. If a class of work carries no images today, a RapidEye capture link opens on any phone, no app to install, for a fixed set of angles taken before the technician sets the record complete.
- Add RapidEye as a user. In Maximo Application Suite, from Administration, with a security group that grants read access to the work order and inspection applications. The connect panel has the details.
What a Maximo work order carries, and who ever looks at it
Checked against IBM's product and documentation pages, September 2026.
Maximo is enterprise asset management: the work, the asset history and the maintenance process for an estate that can run to hundreds of thousands of records. The status field is the spine of that. According to IBM's documentation (ibm.com), the status of a work order indicates its position in the processing cycle and determines which actions can be performed on it, and the documented codes read as a life story of the job.
WAPPRWaiting on approvalThe work order is waiting for approval.APPRApprovedThe work order is approved and the work can begin.WSCHWaiting to be scheduledThe work order is waiting to be scheduled.WMATLWaiting on materialMaterials must arrive before the work can be performed.WPCONDWaiting on plant conditionWork can proceed only when plant conditions suit execution.INPRGIn progressThe work is in progress.COMPCompletedThe physical work is completed. The images taken during the job are now history.CLOSEClosedThe work order is closed and inventory reservations are removed.CANCanceledThe work order is canceled.HISTEDITEdited in historyThe work order was edited in history.Nine of these ten statuses are about whether work can proceed. None of them is about what the asset or the space looked like when the technician walked away, and the two highlighted are where the images stop being looked at.
Where the images actually sit
In two places, and it matters which. IBM's product pages for Maximo Application Suite describe technicians and inspectors conducting inspections on mobile devices and capturing images, videos, forms and operational data as they go, and IBM says that with Maximo Mobile for EAM technicians can complete work orders and inspections with minimal fuss. IBM's guidance on the mobile Inspection application shows inspection forms attached to a work order in the job details and to individual tasks in the plan, questions answered with Pass or Fail, a meter reading or comments, photos or files attached alongside, and the inspection result synchronising back to Maximo automatically. So a photo can belong to the work order as a whole, or to the single question that flagged something. Both are evidence about condition, and both are read.
Can Maximo analyze the photos on a work order?
Yes, through Maximo Visual Inspection, and it is worth understanding exactly what that is. IBM documents that an administrator adds a file upload response to an inspection question and selects the name of the model that Maximo Visual Inspection should use to analyze the photo; during the inspection an inspector can upload a photo and select Analyze photo, continue working while it runs, then review and edit the results, which are presented as a polygon surrounding the identified object. The supported model families include Faster R-CNN, YOLO v3, Detectron2 and an anomaly-optimised option. IBM is clear that this needs both Maximo Manage and Maximo Visual Inspection, and that a model suited to the inspection has to exist; if none matches, the MVI administrator creates one.
That is a precise instrument, pointed at a known defect on a known question by someone who decided in advance what to look for. RapidEye is the wide pass that sits underneath it: it reads every image on every completed record across the estate, including the ones nobody configured a question for, and compares each asset and each space with how that same asset and space have looked before. The two answer different questions, which is why running both is a reasonable thing to do.
IBM Maximo integration FAQ
Does RapidEye integrate with IBM Maximo?
Yes. RapidEye reads the images attached to your Maximo work orders and inspection forms, along with the asset, the location and the time, and returns a condition finding for each issue it sees. Work orders, statuses, assets, job plans and PM schedules in Maximo are not changed.
Can Maximo analyze the photos on a work order?
Yes, through Maximo Visual Inspection. IBM documents that an administrator adds a file upload response to an inspection question and selects the name of the model Maximo Visual Inspection should use, and that during an inspection an inspector can upload a photo and select Analyze photo, with results presented as a polygon around the identified object. It requires both Maximo Manage and Maximo Visual Inspection, and a model suited to that inspection. RapidEye is the wider, per-question-free pass: it reads every image on every closed record across your estate and compares each asset and space with how it has looked before.
What happens to images on work orders that are already closed?
They get read. On first connection RapidEye reviews the image history already sitting on your Maximo records, which is how it learns what each asset and each space normally looks like. A stain that spread over four quarterly PMs is visible in that history long before anyone raised a work order about it.
Will RapidEye change work orders or statuses in Maximo?
No. Maximo stays the system of record. RapidEye reads and reports back, and your planner decides whether a finding becomes a follow-up work order, a defect, or nothing at all.
Does this work across many buildings and sites?
Portfolio scale is the point of it. Findings carry the asset and the location they came from, so a facilities director can see condition rolled up by site, by building or by asset class instead of opening records one at a time.
Do technicians have to work differently?
No. They keep completing work orders and inspections in Maximo Mobile the way they do now, attaching the photos they already attach. Nothing is added to the technician's job.
Other maintenance systems RapidEye reads
Work order and asset systems where the photos land in the same place: on a record somebody closed.
MaintainXPhotos on work orders and on the checklists and inspections that close them
UpKeepWork order and preventive maintenance photos read against the asset
eMaintPhotos captured in the mobile technician experience, read on completion
Brightly Asset EssentialsCondition across school, government and healthcare estates
LimblePhotos from the requester and from the technician who closed the job
All facilities integrations
Sources
- IBM product and documentation pages for Maximo Application Suite: work order statuses, mobile inspections, and photo analysis by Maximo Visual Inspectionibm.com
Sources are named at the publisher level and were checked against the publisher's live pages before this page was published.

