Maintenance log software that turns rounds, readings and repairs into one queryable asset history and statutory registers.

The log becomes something people query rather than something they file, and the readings that predict a failure sit in a column you can trend instead of a book you can only read.
When the inspector asks, the register and the evidence behind it are the same record, and the overdue examination was visible weeks before it became a finding.
No, and it is worth being precise about that. IndeForms does not ship a spare-parts inventory, a purchasing module or an asset-hierarchy master. What it provides is the capture, scheduling, validation, approval, retention and analytics layer around maintenance records, which is the part most CMMS deployments end up doing on paper anyway. Many customers keep their existing CMMS or ERP as the system of record and use connectors, webhooks or the API question type to move data both ways.
No. The asset identity is a field on the form, and there are three ways to get it right: scan the asset tag with a barcode-scanner question, select it from an asset list uploaded as a source so the identifier is picked rather than typed, or pull it live from your own asset master through an API question. Uploaded sources are versioned and cached on the device, so selection works offline. History is then assembled by querying that identifier across submissions.
They answer different questions and we keep them as separate forms. An asset inspection is a condition assessment at a point in time, producing a pass, fail or grade. A work order is a single dispatched job with a completion record, which ends when the job is accepted. A maintenance log is the running record against the asset that accumulates all of it, rounds, readings, repairs and statutory examinations, over the asset’s life. Workflow nodes link them: a failed inspection can spawn a work order, and a completed work order can write its outcome into the asset history.
Yes. The mobile app is offline-first. The form, its asset and reference lists and the compiled rule logic are downloaded to the device before the round, and the completed entry including photographs and audio notes is stored locally and uploaded when the device reconnects. Basements, tank farms and remote pump stations do not cost you a record.
Several controls, used together. A geo control question records where the entry was made, per-field timestamps and geostamps can be enabled at dispatch so each individual answer carries its own time and position, and a geo-proximity rule can warn or block when the position is outside an expected radius of the asset. Because each round exists as a dispatch with its own due window, a late submission is visible as lateness rather than hidden by a handwritten date.
You build the register as a form whose fields match the prescribed ones, with the mandatory fields enforced on the device and a named approver on the record, so an entry cannot be submitted incomplete. The register itself is a filtered export, XLSX or CSV with a column per field, or a PDF, produced on demand or emailed on a schedule. Because submissions stay bound to the published form version they were captured against, the register as it stood on a past date is reconstructable.
Retention is configurable per customer and enforced by a scheduled purge rather than left to habit, so records are held for the period you are required to hold them and not indefinitely by accident. Access is governed by component-level role-based access control, each customer’s data sits in its own database, and the platform can be deployed on-premise or fully air-gapped. Mobile devices can additionally be placed under per-customer data-at-rest encryption with a PIN-derived key and hardware-backed storage.
You can trend any captured reading: submissions are converted to Parquet and queried in-process with DuckDB, so a year of vibration or temperature readings for one asset is a chart, a dashboard widget or a scheduled report. In addition, statistical baselines are computed per field and flag values that are outliers against that field’s own history using z-score, interquartile range and rare-value tests. That is anomaly detection on your captured data, not a physics-based failure prediction model, and we would rather describe it accurately.