In May 2025, Ford halted Explorer SUV production for a week at its Chicago plant. The cause was neither semiconductors nor batteries. It was two heavy rare earth elements: dysprosium and terbium. This came immediately after China's Ministry of Commerce added seven categories of items related to samarium, gadolinium, terbium, dysprosium, lutetium, scandium, and yttrium to its export control list on April 4 of that year, introducing a licensing system. Suzuki halted production of its flagship model, the Swift, starting May 26, and India's Bajaj Auto warned of serious impacts on EV production. According to CLEPA, the European automotive suppliers association, only about a quarter of license applications were approved, and production lines at multiple supplier plants came to a stop.
This turmoil exposed a structural vulnerability in electric vehicle (EV) drive motors. Most high-performance EV motors use sintered neodymium iron boron (NdFeB) magnets, with dysprosium or terbium added to maintain demagnetization resistance in high-temperature environments. That supply chain is extremely concentrated in China. According to a 2025 estimate by research firm IDTechEx, China accounts for 69% of rare earth mining, and when downstream separation, metallization, and magnet manufacturing are included, that share reaches approximately 90%.
How Half a Century of "Conventional Wisdom" Was Formed
The use of rare earth magnets in EV motors was less an inevitability than the result of historical convergence. When NdFeB magnets were commercialized in the 1980s, their high magnetic energy product (, a metric indicating the density of magnetic energy a magnet can store) enabled compact, high-output motor designs. In automotive applications especially, where large output must be packed into limited space, alternatives to NdFeB were rarely considered.
Meanwhile, the concept of the axial flux motor itself is not new. The disk-type generator Faraday built in 1831 was the first axial flux machine, and Tesla patented a disk-type motor in 1889. In this configuration, where magnetic flux flows parallel to the rotation axis, there is a theoretical advantage in power density per volume over radial flux motors. One comparative study reported that, at the same power rating, an axial flux motor had 68.8% higher power density and was 40.6% smaller in volume than a radial flux motor.
However, due to manufacturing difficulties in precisely maintaining tiny air gaps in large disk-shaped machines, axial flux motors long remained confined to the laboratory. The situation changed once rare earth permanent magnets made it possible to secure performance even with larger air gaps, and more recently, precision machining technology has lowered manufacturing barriers.
YASA Demonstrates the "Potential of Axial Flux" at 59 kW/kg
YASA, headquartered in Oxford, UK, became a wholly owned subsidiary of Mercedes-Benz Group in 2021 and has been advancing mass production of axial flux motors. On October 22, 2025, the company announced it had extracted a peak output of 750 kW (approximately 1,000 horsepower) from a 12.7 kg prototype motor, achieving an unofficial world record power density of 59 kW/kg. This represents roughly a 40% improvement over the previous record of 42 kW/kg (550 kW at 13.1 kg) set that same summer. Continuous output is estimated at 350–400 kW.
In June 2026, Mercedes-Benz began mass production of axial flux motors at its Marienfelde plant in Berlin. Fitted to the Mercedes-AMG GT 4-Door Coupe, this motor is under 9 cm wide for the front axle unit and about 8 cm for the rear axle unit, with three units per axle integrated into a single assembly with planetary gears. The front axle motor's rotation speed exceeds 15,000 rpm. For mass production, 35 processes were introduced for the first time in the world, and more than 30 patents were filed.
| Metric | Radial Flux Motor (Typical) | YASA Axial Flux Motor |
|---|---|---|
| Power density | Approx. 5–15 kW/kg (typical range for automotive NdFeB motors) | 59 kW/kg (peak, measured October 2025) |
| Structure | Cylindrical, flux flows radially | Disk-shaped, flux flows axially |
| Rotor arrangement | Inner rotor is common | Two rotors sandwich the stator |
| Manufacturing challenges | Mature mass-production technology | Requires air gap precision, dedicated processes (35 processes are world-firsts) |
| Rare earth use | NdFeB + heavy rare earths (Dy/Tb) is standard | Currently uses NdFeB; aims to eliminate via Project Resilience |
The "Two Paths" Charted by Project Resilience
On August 10, 2026, YASA announced it had secured a grant from the UK government's DRIVE35 program and would launch Project Resilience. DRIVE35 (Driving Research & Investment in Vehicle Electrification) is the core program of the UK's advanced manufacturing sector plan launched in July 2025, allocating £4 billion in grants for automotive R&D, scale-up, and transformation through 2035. Project Resilience is one of 10 projects selected in the program's "Demonstrate" competition, with the 10 projects together receiving £9 million in government funding, and combined government and industry investment exceeding £18 million. YASA's individual grant amount has not been disclosed.
The core of the project lies in developing two rotor technology paths built on a common stator technology foundation.
The first path is the development of permanent magnet motor designs that eliminate heavy rare earths (dysprosium, terbium) for high-performance battery EVs and hybrid vehicles. Magnets themselves are still used, but by removing heavy rare earths as additive elements, the aim is to avoid increasing motor mass or package size while maintaining power density. The APC's project overview explicitly states the goal of "removing dysprosium/terbium from the bill of materials."
The second path is a fully rare-earth-free axial flux technology, addressing vehicle applications with different performance, cost, and manufacturing requirements. Which magnet material this path will adopt has not yet been disclosed.
Tom Hillman, YASA's Head of Motor Simulation, said: "Different applications require different balances of performance, efficiency, packaging, cost, and production volume. That's why we're developing two complementary technology paths rather than a single one-size-fits-all solution."
Under APC requirements, projects funded through the Demonstrate competition must build a product or process demonstrator within 12 months. This means some form of physical verification will be required by summer 2027.
Competing "Rare-Earth-Free" Approaches
YASA's effort is not an isolated one. Efforts to realize rare-earth-free EV motors have accelerated since the 2025 supply crisis.
The most proven approach is the wound-field synchronous motor (EESM), which uses copper electromagnets in the rotor instead of permanent magnets, using no rare earths at all. Renault, BMW, and Nissan have already adopted this for EVs. However, losses from excitation current cause efficiency to drop several points compared to permanent magnet motors, and motor volume tends to increase.
Another approach uses ferrite magnets. The raw materials are widely available and not dependent on China, but magnetic performance falls far short of NdFeB. The magnet volume needed to achieve the same output increases substantially, enlarging the overall motor. IDTechEx analyzes that ferrite motors could be viable for lower-performance vehicle segments.
An emerging new material drawing attention is iron nitride () magnets. Niron Magnetics of Minneapolis, USA, is advancing commercialization, having opened a pilot plant in 2024 and announced joint development with Stellantis in October 2025. On August 7, 2026, the company announced it had secured a conditional direct loan of up to $150 million (20-year repayment term) from the US War Department's (formerly Department of Defense) Office of Strategic Capital. A plant with annual capacity of 1,500 tons under construction in Sartell, Minnesota is scheduled to begin operations in 2027, using only iron and nitrogen—both abundant elements.
| Approach | Rare Earth Use | Power Density Trend | Adoption Track Record | Challenges |
|---|---|---|---|---|
| NdFeB + heavy rare earths (current mainstream) | Yes (Nd, Dy/Tb) | Highest | Nearly all high-performance EVs | China dependence, price volatility |
| Heavy-rare-earth-free permanent magnet (YASA Path 1) | Light rare earths (Nd) only | Aims to match current levels | In development | Ensuring high-temperature resistance |
| Fully rare-earth-free (YASA Path 2) | None | Undisclosed | In development | Magnet material and design undetermined |
| Wound-field synchronous motor (EESM) | None | Tends lower than permanent magnet type | Renault, BMW, Nissan | Excitation losses, increased volume |
| Ferrite magnets | None | Significantly lower | Some low-output applications | Motor enlargement |
| Iron nitride magnets (Niron Magnetics) | None | Potentially close to NdFeB | Pilot stage, targeting 2027 mass production | Coercivity and phase stability challenges |
How "Magnet Geopolitics" Is Reshaping Design Philosophy
What makes Project Resilience interesting is that it aims not merely to substitute materials, but to redesign the motor architecture itself as a "de-rare-earthed" platform. The modular design, in which different rotor technologies can be swapped onto a common stator, gives flexibility to select the optimal magnet strategy for each vehicle program—heavy-rare-earth-free permanent magnets for high-performance models, fully rare-earth-free rotors for cost-focused models.
IDTechEx predicts that by 2036, roughly 30% of the EV market will use rare-earth-free motors. However, this figure is dragged down by China's dominance of both the EV market and the rare earth market, and the proportion of rare-earth-free alternatives is expected to be higher in Europe and the US.
A 12-Month Deadline, and an Answer Not Yet in Sight
The greatest uncertainty facing Project Resilience is whether magnets with heavy rare earths removed can maintain sufficient demagnetization resistance under the high-temperature, high-load conditions of automotive environments. Dysprosium and terbium are added to boost the coercivity (resistance to external magnetic fields) of NdFeB magnets, and removing them raises the risk of demagnetization at high temperatures. YASA's axial flux structure may work thermally in its favor, but no verification data has yet emerged.
Which material will be used in the fully rare-earth-free path also remains undisclosed—iron nitride, ferrite, or perhaps some as-yet-unannounced new material. Whether a demonstrator can be presented within 12 months, as required by the APC, will be the first checkpoint for verification.
Supply security, reduced environmental impact, and maintained power density—no one has yet demonstrated a solution satisfying all three at mass-production scale. Project Resilience's 12 months will be the first to reveal its outline.