Make
maintenance
predictable.

Trusted in research & industry

Trusted by 

World-Class Teams

Detect failures
at the source. Predict them months in advance.

Piemacs is the first integrated package that reads structural strain directly inside the machine — not on its surface. Combined with embedded AI, it turns raw physics into actionable predictions your maintenance team can trust.

See how Mirage turns ideas into stunning digital realities.

0+

International projects

0+

International projects

0

Patented technologies

0

Patented technologies

0

European research programs

0

European research programs

0M

CHF in funding & grants

0M

CHF in funding & grants

The problem

Why most predictive maintenance still fails.

Today's systems chase symptoms — not causes. By the time a vibration sensor triggers an alert, the damage is already advanced.

Failure starts inside

Structural degradation begins at the micro-level — inside the component, invisible to external sensors placed on the housing.

Vibration is a secondary
measurement

Conventional sensors measure a consequence, not the cause. The signal is noisy, attenuated, and arrives too late.

Early signals are ignored

Cascade failures build undetected for months. When the alert finally fires, you're already in emergency maintenance mode.

Cascade failures build undetected for months. When the alert finally fires, you're already in emergency maintenance mode.

AI alone is not enough

Feed a predictive model with the wrong signal and you get wrong predictions. Garbage in, false alerts out.

The solution

One system. Hardware and AI,
end to end.

One system. Hardware and AI,
end to end.

One system. Hardware
and AI,
end to end.

Piemacs combines a patented zero-mass MEMS sensor with embedded AI — in oneintegrated package engineered for harsh industrial environments.

See how Mirage turns ideas into stunning digital realities.

How it works

From metal to insight, in one chain.

From metal to insight, in one chain.

From metal to insight, in one chain.

Source-level signals captured by our hardware, processed by our AI, delivered as actionable predictions your team can act on — performance guaranteed.

Source-level signals captured by our hardware, processed by our AI, delivered as actionable predictions your team can act on — performance guaranteed.

1

Sense

Zero-mass, thin-film MEMS sensor detects strain signals directly inside the machine — at the mechanical source of failure, not the surface.

1

Sense

Zero-mass, thin-film MEMS sensor detects strain signals directly inside the machine — at the mechanical source of failure, not the surface.

2

Process at the edge

On-metal hardware processes raw signals in real time. Edge computing embedded in the sensor.

2

Process at the edge

On-metal hardware processes raw signals in real time. Edge computing embedded in the sensor.

3

Analyze with AI

Physics-grounded AI models — trained on real strain physics, not generic vibration data — identify failure patterns .

3

Analyze with AI

Physics-grounded AI models — trained on real strain physics, not generic vibration data — identify failure patterns .

4

Decide

Prioritized alerts with confidence levels, delivered to your dashboard or CMMS.

4

Decide

Prioritized alerts with confidence levels, delivered to your dashboard or CMMS.

Applications

Built for assets where failure is expensive.

Wherever an unplanned stop costs more than prevention, Piemacs is the bright fit.

Railway wayside

Axle & bearing health

Electric motors

Rotor & winding faults

Robotic joints

Precision arm fatigue

Pumps & pipes

Pressure anomalies

Wind turbines

Gearbox & blade health

Gearbox & blade health

HV transformers

Grid asset protection

Hardware and AI both sides protected.

Two patented layers. One integrated solution. Designed to work together — impossible to replicate separately.

The system

Hardware

The patented sensor

Thin-film MEMS on metal

Shadow installation

Fully customizable to the asset

Resistant to harsh environments

Wi-Fi enabled, low-power

Software

The embedded AI engine

Physics-grounded AI models

Real-time edge processing

RUL forecasting

Prioritized fault detection

Cloud dashboard + open API

Network

Edge + Cloud connectivity

Edge processing on the asset

Secure cloud synchronization

Cloud dashboard + API access

CMMS integration ready

Scalable across your entire fleet

Our team

Built by people who know the physics.

Built by people who know the physics.

Piemacs was born from a real industrial challenge. A client needed a sensor small enough, robust enough, and sensitive enough to detect motor failure before it happens.

See how Mirage turns ideas into stunning digital realities.

Dr. Ramin Matloub

CEO & Co-Founder

MEMS physicist. Serial innovator. EPFL PhD.

Dr. Ramin Matloub

CEO & Co-Founder

MEMS physicist. Serial innovator. EPFL PhD.

Prof. Paul Muralt

Co-Founder & Advisor

EPFL Full Professor. Pioneer in piezoelectric thin films.

Prof. Paul Muralt

Co-Founder & Advisor

EPFL Full Professor. Pioneer in piezoelectric thin films.

Dr. Simone Ngatchou

Lead Engineer

AI & signal processing expert. Industrial systems.

Dr. Simone Ngatchou

Lead Engineer

AI & signal processing expert. Industrial systems.

Backed by physics. Defined by AI. Trusted by Europe.

Trusted in research & industry

Get in touch

Ready
to make failure predictable?

Tell us about your machines, your setup, and your constraints. We'll figure out the right approach together.