Question #260

Can buildings explain how they use energy?

Quinn LabsQuinn Labs
← Work

Project Echo · Research Prototype

Machines are always talking.

Project Echo explores whether sound can help people understand the health of vehicles, equipment and infrastructure before failures become obvious.

Microphone array recording an electric motor on a steel laboratory bench, with a blue waveform traced through the darkness
Concept image. A machine recorded against its own history, not a library of known faults.

What if every machine knew what it was supposed to sound like?

Prototype
Explore Project Echo

The question

What if every machine knew what it was supposed to sound like?

Project Echo explores whether vehicles, machines and everyday equipment can build an acoustic fingerprint of their own health. Instead of waiting for something to fail, could subtle changes in sound reveal problems much earlier?

Can sound become an early warning system for the physical world?

Echo does not try to identify what a machine is, or name a fault from a library. It asks a smaller, harder question: has this machine changed?

Research areas

Vehicle health

  • Cars
  • Vans
  • HGVs
  • Electric vehicles

Audio systems

  • Car speakers
  • Home audio
  • Headphones
  • Musical equipment

Electric motors

  • Fans
  • Pumps
  • Compressors
  • Industrial motors

Buildings

  • Heating
  • Ventilation
  • Water pumps
  • Generators

Infrastructure

  • Rail
  • Factories
  • Renewable energy
  • Public assets

Concept experience

Imagine pointing your phone towards a machine.

Instead of identifying music, it compares today's sound with the machine's own historical fingerprint.

“Rear wheel bearing noise has gradually increased over six months.”

“Left door speaker response is significantly lower than previous recordings.”

“Cooling fan vibration has changed.”

Illustrative research concepts. Not current capabilities.

The sound timeline

Every recording becomes part of a historical record.

  1. 01New
  2. 02Healthy
  3. 03Small change
  4. 04Monitor
  5. 05Inspection recommended
  6. 06Maintenance completed
  7. 07Healthy again

Future research

Exploration only. Not current capabilities.

Acoustic fingerprintsEdge AIPredictive maintenanceVehicle sound analysisSpeaker diagnosticsIndustrial monitoringEnvironmental sound analysisDigital maintenance records

Why it matters

Machines rarely fail instantly.

Most change gradually.

Project Echo explores whether those changes can be recognised earlier through sound, helping people understand equipment before problems become more serious.

Research card

Research PrototypeEngineering • Mobility • AI • Research

Can every machine develop an acoustic fingerprint?

Future possibilities

Research only. None of these exist today.

EV battery acoustic health
Smart maintenance for fleets
Acoustic diagnostics for home appliances
Railway infrastructure monitoring
Wind turbine monitoring
Agricultural machinery
Marine engines
Musical instrument diagnostics
Building services monitoring

Every machine tells a story.

Project Echo explores whether we simply need to learn how to listen.

Talk to Quinn Labs

Research question

“Can every machine develop an acoustic fingerprint?”

Why this matters

Machines rarely fail instantly. Most change gradually, and the change is audible long before it is obvious.

Current progress

Research prototype — recording method and fingerprint stability under test on motors and vehicles.

Workflow

Prototype

Repeat recordings of the same machine, compared against its own fingerprint.

How it could work

Problem

A machine degrades gradually and nobody hears it until it fails.

Current world

Fixed service intervals, or a person who happens to notice a noise.

Quinn Labs idea

Record the machine regularly and compare it with its own past, not with a catalogue of faults.

Potential future

Equipment that raises its hand early, in plain language.

Machine recordedFingerprint storedNew recording comparedChange described in one lineMaintenance logged and fingerprint reset

Research board

Test

Same phone, same spot

Position and microphone matter more than model accuracy. Repeatability is the whole experiment.

Note

Ordinary variation

Cold starts, load, weather. Most change is meaningless — separating it is the hard part.

Reading

Condition monitoring

Industrial vibration monitoring is decades old. Untested at phone-in-pocket scale.

Open questions

  • Changed, not broken

    The question is not what fault this is. It is whether this machine has changed.

See the bench in the Workshop

Future possibilities

Research directions only. None of these exist, and some never will.

EV battery acoustic healthSmart maintenance for fleetsAcoustic diagnostics for home appliancesRailway infrastructure monitoringWind turbine monitoringMusical instrument diagnostics

Questions we're asking

  • Has this machine changed?
  • When did the change start?
  • Is it getting worse?
  • Is it worth inspecting yet?

Potential collaborators

Fleet operatorsMaintenance engineersAcoustics researchersManufacturers

We prototype possibilities. Not every question has an answer. Every answer begins with a question.