# Acoustic cameras find and size compressed air leaks while the line runs, so maintenance fixes the costly ones first (Schneider Electric, Altex/Toyota Industries)

> Acoustic cameras for compressed air leaks: Schneider Electric found 250+ leaks at six sites, saving over $60,000 a year; Altex about 3 million yen.

- Canonical URL: https://aiagentsformanufacturing.com/use-cases/acoustic-camera-compressed-air-leak-detection-sizing-schneider-electric-altex/
- Department: Maintenance (https://aiagentsformanufacturing.com/departments/maintenance/)
- Evidence level: Verified. Real deployment at a named or precisely described company, with a measured result from a primary source.
- Added: 2026-11-12
- Last updated: 2026-11-12
- Publisher: AI Agents for Manufacturing (https://aiagentsformanufacturing.com)
- Manufacturers named: Schneider Electric (https://aiagentsformanufacturing.com/companies/#schneider-electric), Altex (https://aiagentsformanufacturing.com/companies/#altex)
- AI tools deployed: FLIR Si124 (https://aiagentsformanufacturing.com/tools/#flir-si124), CRY SOUND 2620 (https://aiagentsformanufacturing.com/tools/#cry-sound-2620)

## Role

Maintenance and plant energy (compressed air, vacuum and gas leak programs). Typical users: maintenance manager, maintenance technician, plant energy or utilities engineer, TPM or lean coordinator, EHS and sustainability lead

## The problem

Compressed air is one of the most expensive utilities in a plant, and leaks waste a large share of it without anyone noticing: they hiss under the noise of the shop floor, sit in overhead pipes or inside machines, and the classic methods (soapy water, listening, handheld ultrasonic probes) are slow, often need the equipment stopped and do not say how much each leak costs. Schneider Electric reports that compressed air is 1 to 15% of electricity use at its US sites and that leaks can be 20 to 30% of a compressor's output; its teams had trouble convincing stakeholders that leaks mattered and found that ultrasonic probes could not always locate the exact leak. At Altex (Toyota Industries group, aluminium die-cast parts for automotive compressors), leak checks with listening rods or soapy water required stopping equipment.

## How AI helps

An acoustic imaging camera combines a regular camera with an array of dozens of microphones (64 at Schneider Electric's supplier, 124 in the FLIR Si124). Signal processing filters out background noise and overlays a live heat map of ultrasonic sound on the image, so a technician walking the line sees where each leak is, even during production. The camera or its companion software then estimates the size of each leak and its yearly cost, and builds a report, which lets maintenance rank repairs by money lost instead of fixing whatever was found first. The repair itself, and the decision of what to fix, stays with the maintenance team.

## Step by step

1. Measure the starting point: compressor run hours, load and unload time, kWh and, if you can, flow on a weekend when production is stopped (the leak load).
2. Rent or borrow an acoustic camera for a few weeks before buying, and survey one area during normal production: assembly lines with pneumatic tools, overhead headers, quick couplings, FRL units.
3. Tag every leak with an ID, a photo, the estimated flow and cost, and create a work order in your CMMS.
4. Repair in order of cost per year and ease of repair; many leaks are cheap fittings or hoses.
5. Measure again with the same method after repairs, and lower the compressor pressure set point if the network allows it.
6. Make it a routine: a survey every quarter or at each TPM cycle, with results posted where operators can see them.

## Tools

- Acoustic imaging cameras with leak quantification software (64-microphone imagers from a third-party evaluation): used at Schneider Electric
- CRY SOUND 2620 acoustic camera: used at Altex (Toyota Industries group)
- FLIR Si124 acoustic imaging camera with FLIR Thermal Studio and Acoustic Camera Viewer: used at an unnamed automotive parts manufacturer
- Other tools to evaluate: acoustic imagers from Fluke, Sorama, SDT or other suppliers, a leak survey by your compressor supplier, your CMMS to track leak tags and repairs
- To plan the program without new hardware: your compressor data, a leak tag list and any LLM (ChatGPT, Claude, Copilot, Gemini) with the prompt below

## Result

Schneider Electric (US Department of Energy Better Plants implementation model): 40 staff trained on acoustic imagers; more than 250 leaks identified across six facilities, over $60,000 in annual savings and more than 300 metric tons of CO2 avoided per year once repaired; the company plans to use the same tools for gas, steam and vacuum leaks, pump cavitation, noisy bearings and electrical discharge. Altex, Toyota Industries group (article by the camera distributor AirMore): many previously unknown small leaks found in the first TPM activity, 5 to 6 repaired, inspection time much shorter, some machine faults traced to pressure drops caused by leaks; estimated annual savings of about 3 million yen. Unnamed automotive parts manufacturer (FLIR case study, four-week trial): 68 leaks found; fixing only 15 of them cut compressed air use by 54.6 m3 per hour (15%), 230.4 kg of CO2 per day and more than $4,000 a year in electricity.

As reported by the source; most figures are vendor or self-reported and not independently verified.

## Human review: cautions

- The 'AI' here is mostly microphone-array signal processing and automatic leak sizing, not a learning model; the value comes from finding and ranking leaks faster.
- Leak flows and savings from the camera are estimates; check the total with compressor data before and after the repairs.
- Altex's figures come from the camera distributor and FLIR's case is from the camera maker; the automotive parts manufacturer is not named.
- Savings only appear if someone repairs the leaks: plan the repair time, parts and follow-up from the start.

## Try it yourself (prompt)

```text
You are a plant energy and maintenance engineer. I want a ranked plan to fix compressed air leaks.

I will paste: (1) our leak list from a survey [leak ID, location, equipment, estimated flow (l/min or cfm), photo note, likely part to replace]; (2) our compressor data [pressure set point, kW, run hours per year, electricity price per kWh].

Do this:
1. Estimate the yearly energy cost of each leak from its flow and our compressor data, and show the formula you used.
2. Rank the leaks by cost per year and by ease of repair (fitting, hose, coupling, valve, needs a shutdown), and group the ones that can be fixed in one round.
3. Draft a work order text for the top ten (location, part, tools, whether a stop is needed).
4. Give the total expected saving if the top ten are fixed, as a range, and say what would make it wrong.
5. Suggest how to check the saving afterwards with the data we already have.

Do not invent flows or prices: if a value is missing, mark it. Use only the data I paste in; if something is missing, write 'not found' or list it as a question instead of guessing. Do not send, order, post or change anything in any system: produce drafts for my review and approval.
```

## Related use cases

- [A walking robot runs thermal and acoustic inspection rounds and turns what it finds into work orders (AB InBev, BMW, Nestlé Purina)](https://aiagentsformanufacturing.com/use-cases/robot-thermal-acoustic-inspection-rounds-ab-inbev-bmw/)
- [AI cuts plant energy cost: demand-based ventilation control, price-aware production plans and cross-plant energy benchmarking (Stellantis, Casuso Propellers, Jotun)](https://aiagentsformanufacturing.com/use-cases/ai-plant-energy-optimization-ventilation-production-planning-stellantis-casuso-jotun/)
- [A camera or a phone photo reads the analog gauges, so inspection rounds are no longer written by hand (Nachi-Fujikoshi, Kobe Steel group, a plastics maker)](https://aiagentsformanufacturing.com/use-cases/ai-analog-gauge-reading-camera-inspection-rounds-nachi-kobe-steel/)
- [AI reads oil condition from sight-glass photos and lab reports, so lubricant changes follow the machine, not the calendar (vacuum pump maker, railcar component maker)](https://aiagentsformanufacturing.com/use-cases/ai-oil-condition-monitoring-lubrication-analysis-vacuum-pumps-gearboxes/)

## Sources

- https://betterbuildingssolutioncenter.energy.gov/implementation-models/schneider-electric-using-state-art-acoustical-imaging-technology-compressed
- https://n-airmore.com/en/column/interview3
- https://www.flir.com/discover/industrial/case-study-tackling-compressed-air-leaks-in-automotive-parts-manufacturing-with-acoustic-imaging/

---
Please cite the canonical URL: https://aiagentsformanufacturing.com/use-cases/acoustic-camera-compressed-air-leak-detection-sizing-schneider-electric-altex/
