MHIRC Machine Health Intelligence
Industrial AI research · Van, Türkiye

Reading machine health from the signals machines already emit.

We build deep-learning systems that detect, classify and forecast faults in rotating machinery — from distributed bearing damage to inverter failure — using vibration, stray magnetic flux and motor current. No new instrumentation on the shaft, no downtime to measure.

Vibration traces: a healthy bearing above, a distributed fault below showing impulsive bursts g 0 0 0 0.25 0.5 0.75 1 s HEALTHY DISTRIBUTED FAULT
3-axis vibration · x-channel healthy  /  faulted
6 Sensor channels fused per diagnosis
50 Speed and torque operating points
3–10 hp Motor platforms characterised
5 Peer-reviewed IEEE publications

Introduction

What we do, in fifty-one seconds

Two bearings, one of them failing — then the hardware, the signals and the record behind the work.

Narrated in English. The on-screen text carries the same information, so the film also reads with the sound off.

About the centre

A research group built around one question: what is this machine about to do?

The Machine Health Intelligence Research Centre works on condition monitoring, fault diagnosis and prognostics for electrical machines. Our focus is the hard case rather than the textbook one — distributed bearing damage that produces no clean defect frequency, compound faults that mix mechanical and electrical signatures, and models that must hold up when the load, the speed or the measurement session changes.

Real faults, not simulated ones

Our datasets come from bearings degraded by lubrication loss, contamination, electrical erosion and flaking — the failure modes that actually dominate industry, recorded on instrumented testbeds rather than synthesised by notching a race.

Many signals, one decision

Three-axis vibration, stray magnetic flux and phase currents each see a different part of the failure. We fuse them in architectures designed for the fusion itself, and we measure what each channel is actually worth.

Protocols fixed before the result

Cross-load, cross-speed and leave-one-session-out evaluation, with significance testing over many fits. We report where a method fails as carefully as where it works, because a diagnosis that only holds in one session is not a diagnosis.

Research areas

Where we work

Six active lines, from sensing hardware through to deployment-grade generalisation — plus a second stream applying the same signal-intelligence methods to biomedical diagnostics.

Distributed bearing fault diagnosis

Distributed damage spreads energy across the spectrum instead of concentrating it at a defect frequency, which is why classical envelope analysis struggles. We design 1-D, 2-D and 3-D convolutional architectures with hybrid inputs — raw time frames alongside engineered features — that separate lubrication loss, contamination, erosion and flaking from a healthy baseline across the full speed–torque envelope.

1D / 2D / 3D CNN Hybrid inputs Residual blocks 50 operating points

Remaining useful life prognostics

Fitting an accelerometer to a small cooling-fan motor is rarely practical or affordable. We estimate remaining lifespan from three-phase current alone, using residual and channel-attention blocks — in low-rated-torque machines the modulated torque disturbance from a degrading bearing is clearly legible in the phase currents.

Channel attention Current-only PMSM / BLDC

Multi-sensor instrumentation

A custom sensory board integrating three-axis vibration, a fluxgate stray-flux sensor and two-phase current acquisition — plus the study of how far ADC resolution and sampling rate can be reduced before diagnostic accuracy degrades.

Fluxgate 8–16 bit

Domain generalisation and transfer

A model that only works at the speed it was trained on has no industrial value. We use multimodal contrastive learning and order-domain representations to build diagnoses that survive changes in speed, load and acquisition session.

Contrastive Order tracking

Motor and inverter fault diagnosis

Beyond bearings: open-circuit and switch faults in PMSM drive inverters, broken-bar signatures, eccentricity, and the compound case where mechanical and electrical faults are present at once and their signatures interfere.

PMSM Compound faults

Biomedical signal intelligence

A bearing race and a heart are both systems whose condition is encoded in a noisy periodic signal. The same methods carry over: ECG arrhythmia classification, Parkinsonian gait analysis, sleep apnoea detection and musculoskeletal imaging with explainable attribution.

ECG / PTB-XL Gait Grad-CAM

Capabilities

What we can do for a partner

We work with industry and with academic groups. Engagements typically start with a feasibility study on your own data and end with a model you can run on your own hardware.

Condition-monitoring feasibility study

You send representative recordings from your machines. We report which faults are separable from the signals you already have, which extra channel would buy the most, and what accuracy is realistic before anyone commits to a rollout.

Custom diagnostic model development

End-to-end: signal conditioning, feature and representation design, architecture selection, and an evaluation protocol fixed in advance so the reported number is the one you will actually see in the field.

Sensor selection and cost reduction

Minimal feature sets and reduced ADC resolution can cut the cost of a monitoring node substantially. We quantify exactly how much accuracy each reduction costs, so the trade-off is a decision rather than a guess.

Prognostics and maintenance scheduling

Remaining-useful-life models with calibrated uncertainty, so a maintenance window can be planned against a probability rather than a point estimate.

Independent model review

An outside read on an existing diagnostic pipeline: leakage between train and test, session confounds, over-optimistic protocols, and whether the reported accuracy would survive a change of load.

Joint research and training

Co-supervision of graduate work, collaboration on funded projects and grant applications, and short technical courses on signal processing and deep learning for machine health.

Power-electronic converter design and prototyping

Multilevel inverter topologies for drives, electric vehicles and renewable systems — from topology selection and switching strategy through to a working laboratory prototype with measured efficiency and harmonic performance.

Converter and drive-side fault diagnosis

Open-circuit switch faults, dead-time and shoot-through failures in inverters. The same diagnostic discipline we apply to bearings, applied to the converter feeding the machine — so a drive fault is not misread as a mechanical one.

FPGA and embedded implementation

Modulation and control laws realised on FPGA, and diagnostic models compiled down to a node that runs on the machine rather than in a datacentre. What is demonstrated in the laboratory is what gets deployed.

Approach

How a diagnosis gets built

The same five steps, whether the subject is a 10-hp induction motor or a twelve-lead ECG.

Put each signal on the axis where it is stationary

Vibration resampled on shaft angle fixes bearing and eccentricity coordinates. Stator-current signatures are anchored to supply frequency and slip instead, so forcing both onto one axis smears one of them. Getting the representation right is worth more than any architecture choice downstream.

Characterise the full operating envelope

Data across ten speed levels and five torque levels, healthy and faulted, so the model sees the variation it will meet in service rather than a single convenient operating point.

Fuse channels deliberately

Vibration, flux and current are combined in architectures built for multi-input fusion — then ablated, so we know what each modality contributes rather than assuming more sensors is better.

Test the transfer, not just the fit

Change the load. Change the speed. Hold out an entire acquisition session. The gap between those numbers and the in-distribution number is the honest measure of whether a model is deployable.

Reduce until it is affordable

Minimal feature subsets, lower sampling rates, fewer ADC bits, smaller models. A diagnosis that needs a laboratory to run is a result; one that runs on a monitoring node is a product.

Research impact

Work that other researchers build on

The figures below are the Google Scholar record of our founder, Dr. Ramin Rajabioun, as of October 2026. They are his personal publication record rather than the centre’s, and every number can be checked at the source.

2,386 Total citations
11 h-index — eleven papers cited at least eleven times each
1,561 Citations of one algorithm: the Cuckoo Optimization Algorithm
18 yrs Cited every year without interruption since 2009
Citations per year
Citations per year, 2009 to 2026 Bar chart of annual citations. Rising from 9 in 2009 to a peak of 205 in 2016, then holding between 131 and 189 every year since. The 2026 bar is hatched because that year is still in progress. 50 100 150 200 2009: 9 citations 2010: 37 citations 2011: 49 citations 2012: 82 citations 2013: 124 citations 2014: 157 citations 2015: 177 citations 2016: 205 citations 2017: 189 citations 2018: 163 citations 2019: 166 citations 2020: 170 citations 2021: 157 citations 2022: 161 citations 2023: 142 citations 2024: 131 citations 2025: 164 citations 2026: 72 citations (year to date) 205 72 2009 2012 2015 2018 2021 2024 2026 2026 is still in progress

Impact here is sustained rather than a single spike: the count has not dropped below 130 in any complete year since 2014.

Selected publications

Peer-reviewed work

Published in IEEE Transactions, IEEE journals and IEEE conference proceedings.

View the full publication list

Team

Who we are

Dr. Ramin Rajabioun

Founder & Director

B.Sc. in Biomedical Engineering (Sahand University of Technology), M.Sc. in Control Systems Engineering (University of Tehran), Ph.D. in Electrical and Electronic Engineering (Van Yüzüncü Yıl University, 2025). Works on evolutionary computation and optimisation, deep learning, computer vision and large language models, applied to machine health and biomedical diagnostics.

19910001153@yyu.edu.tr

ORCID 0000-0003-3972-6442

Dr. Hasan Hataş

Co-founder

B.Sc., M.Sc. and Ph.D. in Electrical and Electronics Engineering from Van Yüzüncü Yıl University, where he is now an Assistant Professor. He has published widely and worked on research projects developing efficient energy conversion systems. His interests are power electronics, electrical machines, renewable energy systems and the control of electric drives.

hasanhatas@yyu.edu.tr

ORCID 0000-0003-4543-362X

Collaborators

Across three universities

Our published work is joint with researchers at the University of Texas at Dallas, Texas A&M University–Commerce and Van Yüzüncü Yıl University. We are open to new collaborations on condition monitoring, prognostics and applied signal intelligence.

Contact

Start a conversation

If you have machines you would like to monitor, data you would like assessed, or a research collaboration in mind, write to us. A short description of your setup and what you are trying to detect is enough to begin.

Location Van, Türkiye
Areas Condition monitoring · Fault diagnosis · Prognostics · Biomedical AI