← NAOAT News

Percepto opens its inspection analysis platform to third-party data

Axion separates inspection analytics from the choice of drone fleet, with asset history at the center of the company’s software offering.

Conceptual flying inspection robot
AI-generated illustration. Conceptual flying inspection robot. AI-generated conceptual illustration; not a depiction of the named organization or its equipment.

Percepto introduced Axion on October 27 as a software service for infrastructure operators using supported third-party drones, sensors and existing inspection archives. The launch widens the company’s offering beyond customers deploying its own autonomous aircraft.

The company says Axion organizes observations around individual assets, combining inspection history with specialized models and expert validation. Its examples include interpreting a thermal anomaly against expected equipment behavior and identifying a downed wire in relation to nearby components. Operators can also query the stored findings in ordinary language.

Percepto reports more than 160 specialized models and over 100,000 hours of training data from autonomous inspections. These are company figures, not an independent comparison of detection performance. Its announcement also describes a multi-year transmission-inspection contract with an unnamed U.S. energy company; it does not disclose the contract’s value.

The service was announced as available at launch. Percepto’s Ranger aircraft and Harbor stations remain optional collection systems within the expanded offering.

NAOAT analysis

The significant commercial change is the separation of data collection from interpretation. An operator with an established fleet may be able to evaluate new analytics without replacing aircraft. That shifts part of the purchasing decision toward software compatibility, asset records and the quality of findings.

Historical comparison creates a demanding integration problem: a picture from one inspection must be associated with the same component in another. A mistaken association could make an unchanged asset appear to be deteriorating, or obscure a real change. Hardware flexibility therefore increases the importance of consistent identification and provenance.

The launch establishes a product direction. Evidence about missed defects, unnecessary alerts and how expert review affects turnaround would establish its operational value. Model counts and training hours alone cannot answer those questions.

Sources & attribution

Written with AI assistance from the primary source linked below. The analysis section is NAOAT analysis. Source dates are separate from this article’s publication date.

  1. Percepto — Axion launch announcement ↗Source published October 27, 2026