TERMINATOR – Thermal and Electrical Reliability Modeling of INverter-induced Ageing in moTORs

“Yes, the project is called TERMINATOR. Unlike the movie, our goal is not to stop the machines. It is to help them survive longer.”

Project Information

  • Project Type: Ph.D. Project
  • Funding Agency: Swedish Energy Agency (Energimyndigheten), FFI Nollutsläpp Programme
  • Industrial Partners: Traton AB, ABB AB, ABB Robotics Sweden AB
  • Total Budget: 10.5 MSEK
  • Public Funding Requested: 6.0 MSEK
  • Project Duration: October 2026 – August 2030 (47 months)

Why are we doing this?

The electrification of trucks, buses, robots, and industrial machines is accelerating rapidly. Modern electric drives increasingly rely on silicon-carbide (SiC) and gallium-nitride (GaN) power electronics. These technologies make electric vehicles more efficient, more compact, and more powerful.

However, there is a catch.

These modern power electronics switch electrical voltages much faster than previous generations of converters. While this improves efficiency, it also exposes electric motors to new types of electrical stress that may accelerate ageing of insulation systems and bearings. Think of it as replacing a daily jogging routine with an endless sequence of sprints. Performance improves, but long-term wear may increase dramatically.

Today, surprisingly little is known about how these ultra-fast switching technologies affect the lifetime of inverter-fed electric machines. This uncertainty creates challenges for designers of the next generation of electric trucks, industrial drives, and robotic systems.

Figure: Wide-bandgap power electronics such as SiC and GaN enable faster and more efficient electric drive systems, but also introduce new reliability challenges.

Project goal

The goal of TERMINATOR is simple:

Understand how fast-switching inverters affect the ageing of electric machine components and develop tools that can predict failures before they happen. More specifically, the project aims to:

  • Identify degradation mechanisms in motor insulation systems.
  • Study electrical ageing of rolling bearings.
  • Detect measurable indicators that reveal early damage.
  • Develop predictive lifetime models.
  • Validate the results on real electric machines.
  • Improve reliability of future electric powertrains.

How Will We Do It?

1. Build Dedicated Ageing Laboratories

Several experimental test benches will be developed at KTH to recreate the electrical stress
produced by modern SiC and GaN inverters.


Figure 2. Conceptual illustration of the ageing test platforms that will be developed at KTH.

2. Accelerated Ageing Experiments

Motor insulation samples and bearings will be exposed to controlled electrical and thermal
stress for hundreds or thousands of hours.

The idea is similar to compressing decades of operation into a much shorter testing period.

3. Detect Damage Before Failure

Rather than waiting until a component completely fails, experiments will be stopped at
different stages of ageing. This will allow researchers to study what happens during the
early stages of degradation.

4. Develop AI-Assisted Reliability Models

Advanced data analysis, signal processing and machine-learning methods will be used to
discover relationships between:

  • Switching frequency
  • Voltage slew rate (dv/dt)
  • Temperature
  • Electrical measurements
  • Degradation behaviour
INSERT_IMAGE_URL_HERE
Figure 3. Experimental data will be used to generate predictive ageing and reliability models.

5. Validate on Real Machines

Finally, the developed models will be validated on inverter-fed electric machines in the laboratory.
Because if a model only works in a spreadsheet, industry tends to remain unconvinced.
🙂

Results So Far

The project officially starts in October 2026, so experimental results are not yet available.

However, several important achievements have already been completed:

  • A complete experimental methodology has been developed.
  • Key industrial requirements have been identified.
  • A detailed work-package structure has been established.
  • Critical knowledge gaps have been mapped through literature studies.
  • Strong industrial participation has been secured from Traton, ABB and ABB Robotics.

Expected Impact

If successful, TERMINATOR will help answer a fundamental question:

How long will electric motors survive when connected to tomorrow’s ultra-fast power electronics?

The project is expected to contribute to:

  • More reliable electric trucks and buses.
  • Lower maintenance costs.
  • Improved drivetrain design methods.
  • Better understanding of insulation and bearing degradation.
  • More accurate lifetime prediction tools.
  • Stronger Swedish competence in electrified transportation and industrial electrification.

Building on Previous KTH Research

TERMINATOR builds upon several years of research at KTH on electrical machine reliability,
condition monitoring, and insulation ageing. In particular, the project benefits from earlier
research on inverter-fed motors and stator winding condition monitoring.

INSERT_IMAGE_URL_HERE
Figure 4. Previous KTH experimental platform used for insulation ageing studies.

Future blog posts will connect TERMINATOR with other ongoing projects in:

  • Electric machine reliability
  • Condition monitoring
  • Wide-bandgap power electronics
  • Artificial intelligence for diagnostics
  • Electrified heavy transport

Further Reading

  • https://doi.org/10.1109/MPEL.2024.3519924International Technology Roadmap for Wide Bandgap Power Semiconductors
  • https://doi.org/10.1109/TED.2023.3346369Review and Outlook on GaN and SiC Power Devices
  • https://doi.org/10.1109/ACCESS.2022.3225119Review of Bearing Currents in Electrical Machines with Variable-Frequency Drives
  • https://doi.org/10.1109/OJPEL.2021.3069780Partial Discharge Study of Motor Winding Insulation under High dv/dt Pulses

Project Tagline

TERMINATOR studies how tomorrow’s ultra-fast power electronics age electric motors,
helping future electric trucks, robots, and industrial machines stay efficient,
reliable, and alive for longer.