Autonomous inspection robots & drones

Field Control Devices Industry Acceptance Near-term · 2–5 yrs Evidence E3 · First production use AI / ML

Legged and wheeled robots, and drones, that run routine inspection rounds in process plants, including in explosion-hazard zones. Certified hardware exists and named operators run it; turning the data into maintenance work is the hard part.

EMERGINGPROMISEREALITYACCEPTANCEPROVEN
Autonomous inspection robots & drones is placed at the Industry Acceptance stage, with an estimated time to proven use of 2–5 yrs (near-term) and evidence level E3, first production use. The dashed ring marks an AI/ML-driven entry. Stage reflects attention versus results; the evidence level reflects deployment proof. The height of the curve is illustrative only. October 2026 edition.

What the robots do

Operator rounds remain a large part of plant work: walking a route to read local gauges and sight glasses, look and listen for leaks, check pumps and take thermal images. Inspection robots automate the repetitive part. A robot follows a mapped route, stops at defined inspection points, captures visual, thermal and acoustic data, reads analogue gauges with image recognition, and returns to a dock to charge and upload results.

There are three main platform types. Legged robots climb stairs and step over pipes and cable trays, which matters on multi-level process structures. Wheeled and tracked robots are simpler and carry heavier payloads, but need suitable floors and ramps. Drones reach elevated structures such as flare stacks and tank roofs, usually under closer human supervision.

Hazardous areas are the dividing line

Much of a refinery or chemical plant is classified as a potentially explosive atmosphere, and equipment used there needs ATEX or IECEx certification for the zone. An uncertified robot can work only in safe areas or under special permits, which defeats the purpose of routine autonomous rounds. Certified platforms are therefore the key development.

Shell developed an early robot, the ExR-1, with ExRobotics; it was certified for IECEx Zone 1 environments and intended for Shell's remote facilities 3. Its successor, the wheeled ExR-2, is ATEX-certified and was launched in collaboration with Shell Pernis and Dow Terneuzen 4. ANYbotics' ANYmal X is an ATEX-certified legged robot for Zone 1, indoors and outdoors 2. At BASF's Ludwigshafen site it has run multi-floor missions in the steam cracker and the oxo plant, each covering 20 inspection points in about 20 minutes: reading levels, gauges and lever positions, taking thermal images of tanks, pumps and furnace walls, and recording pump acoustics 1.

The BASF case is instructive for what it reports and what it does not. It describes the missions in detail and their potential to trigger preventive maintenance, but gives no measured outcomes 1. That is typical of public information in this field: deployments are real, but published results on cost, finding rates or avoided incidents are scarce.

The real work is integration

A robot that collects thousands of images a week is only useful if someone or something reviews them and the findings reach the maintenance system. The path usually runs from the robot to a fleet-management platform that plans missions and analyses data, then on to the historian and the CMMS. Each link needs work: gauge-reading accuracy must be checked against manual readings, anomaly thresholds tuned, and findings mapped to equipment tags and work-order codes.

Connectivity is the other dependency. Robots upload large volumes of data and may be supervised remotely, so they need reliable wireless coverage along their routes, which is often what triggers site-wide Wi-Fi or private cellular projects. Site procedures change too: emergency response, permits, and how robots and people share walkways all need defining.

What it does not replace

Robots do not replace an operator's judgement or hands-on intervention. Their value is consistency and frequency: the same reading taken the same way every shift, a continuous thermal record, and fewer people sent into hazardous or remote areas for routine checks. A business case built on reduced exposure and better data is easier to defend, and to deliver, than one built on headcount reduction.

Evidence timeline

  1. Mar 2018Shell and ExRobotics present the ExR-1, an IECEx Zone 1-certified inspection robot. 3
  2. Aug 2023ANYbotics reports ANYmal X missions in BASF Ludwigshafen's steam cracker and oxo plant. 1
  3. Nov 2023ExRobotics' ATEX-certified ExR-2, developed with Shell Pernis and Dow Terneuzen, reported in trade press. 4

Why it sits at the Industry Acceptance stage

We have moved this entry earlier within the Industry Acceptance stage in this edition. Certified hardware exists and named operators run it in live plants, which takes it beyond the Reality Check stage. But the previous description of deployment "at scale" overstated the public evidence: published deployments are mostly single sites or limited fleets, and outcome data is thin.

Adoption signals to watch

Questions to ask before committing

Terms used on this page

Autonomous mission
A pre-planned route the robot executes without a remote pilot.
Acoustic imaging
Locating gas leaks or electrical discharge from the ultrasound they emit.
Docking station
The base where a robot charges and transfers data between missions.

Sources

  1. ANYmal X: BASF Ludwigshafen deployment ANYbotics, Aug 2023
  2. ANYmal X ANYbotics, n.d.
  3. Shell robotic inspection for the global energy industry (ExR-1 case study) Materialise, Mar 2018
  4. ExR-2: the autonomous inspection robot for the petrochemical industry Inspenet, Nov 2023

Related technologies

See this entry on the full maturity map · All Control Devices entries · How placements are made. Stage and time-horizon placements are editorial judgements based on the public sources listed above, as of October 2026. They are not engineering, safety or investment advice; see the Terms of Use.