Dual Flux Technology™: a new architecture for electric motor design.

Dual Flux Technology™ is a new electric motor architecture that combines the two normally exclusive methods of generating rotor magnetic field within a single rotor.

Conventional electric motors usually take one of two approaches. Permanent magnet motors (PMSM) fit magnets into the rotor, giving a magnetic field that costs no energy to maintain. Electrically excited motors (EESM) put coils in the rotor and feed them with current, giving a field that can be varied but must be continuously powered. Each approach has an area on the operating map where it performs well and one where it does not.

Dual Flux incorporates both approaches. A reduced quantity of permanent magnet provides field in the low-speed, high-torque area where a magnet is most valuable while electrically excited coils add flux for peak torque demand and reduce it during high-speed cruising. The balance between the two is set at design time to match the duty cycle of the specific application.

The result is a motor that delivers permanent-magnet-class performance with substantially less magnet material and without the efficiency penalty that either architecture pays in isolation.

Dual Flux is supplied as a complete system: the machine itself, an inverter with excitation control, framework field control and dual-flux energy optimisation hardware, and the control software that manages the two flux sources seamlessly.

The difference lies in where the rotor field comes from and what costs arise at different operating areas.

• PMSM: field from permanent magnets fixed in the rotor. Strong at low-to-mid speed, torque density and thermal performance; constrained by high-speed efficiency (the magnet's field must be actively opposed) and by expensive, supply-constrained materials.
• EESM: field from coils in the rotor, supplied with current. Strong at high-speed efficiency, peak torque capability and low-cost materials; constrained by low-speed and urban efficiency (excitation current is paid continuously) and by more complex design and control.
• Induction: field induced in the rotor by slip. Strong on cost and robustness; constrained by efficiency, particularly at part load.
• Dual Flux™: both, within a single rotor. Strong on efficiency across the full speed and torque range and on peak and continuous power capability; the added control complexity is managed within the Talos inverter and software stack.

The practical distinction is easiest to see in a drive cycle. In urban operation, a permanent magnet motor is efficient because its field is free. An electrically excited machine must push current through its rotor coils simply to have a field at all and that current becomes heat. At sustained high speed, the position reverses. The permanent magnet continues generating field whether it is needed or not, and the controller must spend current opposing it; current that produces no torque and is pure loss. An electrically excited machine simply reduces its excitation and runs efficiently.

Dual Flux does not choose. It uses the magnets to cover the low-speed region at no energy cost and reduces excitation directly at high speed rather than fighting the magnet with stator current. Because the magnet is sized for the low-speed region rather than for peak power, magnet mass falls sharply.

There are two consequences that matter to a design team:

• The torque-per-amp optimum becomes a design variable rather than a fixed consequence of the topology. Two applications with different duty cycles can receive two differently balanced motors from the same architecture.
• The magnet operates under less stress. Because field is reduced at source rather than opposed by stator current, the magnet is not driven as hard toward its demagnetisation limit. This is relevant to high-temperature duty cycles, which is normally where permanent magnet designs require additional heavy rare earth content.

One further difference from conventional EESM: dual-flux motors can utilise either slip rings or wireless inductive excitation. With wireless excitation, this removes the brush and slip-ring wear mechanism that has historically made electrically excited rotors unattractive for high-volume consumer products.

Measured on the Talos dual-flux demonstrator under standard test cycles:

• 96.2% peak system efficiency
• 93.1% efficiency over NEDC
• 20,000 rpm maximum operating speed

Programme-specific targets are established through the NeurotorQ design software against the customer's own duty cycle. The drive-cycle figure is the more meaningful of the two. Peak efficiency is a single operating point; efficiency across a complete cycle reflects what a vehicle actually delivers in service and it is where a single-excitation machine gives ground.

Against benchmark comparisons, Dual Flux delivers up to 10-30% reduction in motor energy consumption over a drive cycle. The benefit varies with the cycle. An application spending most of its time in one operating region will see less benefit than one with genuinely mixed duty.

On power density and material content:

• Up to +30% power density compared with standard benchmark powertrains
• Up to −40% weight, derived from benchmark powertrain comparisons
• Up to −80% magnet material, reducing critical supply chain dependency
• Up to −30% cost, compared with alternative solutions at volume

As a practical demonstration, in a 55 kW motor study (our Hipermod project), a permanent magnet rotor required 0.52 kg of magnet, while the Dual Flux rotor achieved comparable performance coverage across city, rural and motorway regions with 0.15 kg, under a third of the magnet mass.

The architecture scales across power classes and is suited to any application where duty cycle is mixed, magnet cost or supply is a concern, and packaging and weight are constrained.

Current focus areas:

• Electric motorcycles: compact, lightweight, high-output systems where cost sensitivity is high, packaging is tight and the duty cycle genuinely spans urban and sustained high-speed running.
• Electric vehicles: passenger and commercial platforms, where range and real-world efficiency drive the value.
• Robotics: precision actuation requiring responsiveness, torque density and thermal headroom.
• Unmanned vehicles (UAVs, USVs, UGVs, UUVs) and eVTOL systems: motors optimised for endurance, weight and flight performance. Talos's first commercial contract was in electric aerospace.

Additional segments served: marine, motorsport, aerospace and industrial applications, typically at lower volumes and often with full system supply rather than technology transfer.

Hybrid applications: a Dual Flux hybrid unit supports idle and moving stop-start, engine load shifting, torque assist and boost, sailing and coasting, energy recuperation and brake regeneration, and removes the need for a separate 12 V starter. For a manufacturer not ready to move fully electric, this offers meaningful efficiency gains on an existing platform.

Yes, to both. Integration into existing chassis and drivetrain systems is an explicit design objective of the architecture and licensing is our primary commercial model.

There are two deployment routes:

• High-volume applications (cars, motorcycles, commercial vehicles, UVs): deployment through design and royalties. Talos designs the motor and transfers the technology; the manufacturer produces at their own scale. The customer ends up owning and building their own implementation rather than taking a supply dependency on Talos.
• Low-volume applications (aerospace, marine, motorsport, industrial): deployment through design, royalties and/or full system supply. Where tooling up is not economic for the customer, Talos can supply the complete system: machine, inverter and controls.

Engagement proceeds in four stages, each proving the case before the next commitment is made:

• Stage 1: Concept Design (no cost). Under a non-binding MOU, Talos funds the first design iteration against the customer's specification, using our proprietary software platform, NeurotorQ.
• Stage 2: POC Virtual. Under an engineering NRE budget, design iterations are executed using our NeurotorQ platform in order to optimise and fully define the specification to prototype-build readiness.
• Stage 3: POC Physical. Under an NRE budget with Talos co-investing, a small number of physical prototypes are built, typically two or three, with detailed test results.
• Stage 4: Technology Transfer. Under a technology agreement, motor and system specifications are adapted to Dual Flux across the platform, with the benefits available at scale.

The first stage costs the customer nothing beyond a non-binding MOU. This is deliberate: NeurotorQ makes concept design fast enough that we would rather demonstrate fit against a real specification than argue for it in a meeting. Whether the architecture fits a specific existing platform (packaging envelope, cooling, mounting, drivetrain interface) is a concept design question rather than a datasheet one and Stage 1 exists to answer it.

Talos Technology is headquartered in Banbury, United Kingdom, and was founded in 2018.

All technology development takes place in the UK at the Banbury Technical Centre, established in 2023. The facility is an electrical machines and powertrain development centre covering the full prototype development cycle in-house:

• Design and analysis: advanced mechanical design, design for manufacture, non-linear mechanical analysis, CFD and thermal analysis, 2D and 3D electromagnetic analysis, dynamic system modelling and production drawings to BS 8888.
• Manufacturing: EDM wire cut for stator and rotor lamination, in-house lamination treatment, stacking and forming, CNC winding machine for fast production scaling, CNC lathe work for shafts and cooling jackets and a 5-axis CNC with full CMM inspection capability.
• Test and validation: an in-house motor test rig rated at 22.5 kW continuous and 33 kW peak, up to 20,000 rpm, peak torque to 275 Nm, with a 30 kW / 500 V / 385 A bidirectional DC supply and three-phase AC and DC power analysis. Specialised cooling validation rigs and our own ABAX test automation software support the programme, alongside partnerships with industry-leading test houses.

The centre also includes a prototype assembly facility with mechanical stations for rotor and stator assembly, automated quality assurance, an electronics development testbench and a fully equipped motorcycle workshop for vehicle modification and build.

NeurotorQ is Talos's in-house electric machine design and validation proprietary software platform. It compresses time spent for coupled electromagnetic, thermal and mechanical exploration and performs full feasibility and performance validation before any physical part is built.

It is the software counterpart to Dual Flux: the architecture defines what is possible and our NeurotorQ software designs a tailored motor for that specific application.

• Risk moves to the front of the programme, where it is cheap. Performance against the customer's own duty cycle is modelled before budget is committed and before metal is cut.
• It is why Stage 1 concept design can be offered at no cost. The speed of the software platform is what makes a free first iteration economically rational for us.
• It is what makes "customizable by design" affordable. Producing differently balanced motors for different duty cycles from one architecture would be prohibitively expensive without it.
• Designs are production-ready. Topology optimisation is driven by electromagnetic parameters rather than structural ones alone and designs are developed to be built with next-generation production equipment.

Standard design software handles electromagnetic, thermal and mechanical domains as separate analyses that an engineer couples manually, which is why exploring a genuinely new topology has historically cost weeks per variant. That cost is the main reason design teams iterate on what they already have rather than questioning the architecture.

NeurotorQ utilises AI functionality in order to explore many possible designs simultaneously, optimising based on client specifications. This enables it to discover designs that are lighter, more efficient and better suited to how the system actually operates.

It runs throughout and it is the explicit engine of Stage 2. Stage 1 concept design produces a first iteration against the customer's specification; Stage 2 (POC Virtual) utilises NeurotorQ in order to execute multiple iterations to optimise and fully define the specification to prototype-build readiness, before any physical prototype is committed in Stage 3.

The architecture allows the use of both slip rings and wireless inductive excitation. In the wireless configuration, excitation is inductive, which removes the brush and slip-ring wear mechanism associated with conventional electrically excited machines. This is one of the principal reasons the architecture is viable for high-volume consumer applications.

It requires excitation control alongside conventional armature field control, plus the energy optimisation hardware that manages both flux sources. Talos supplies the inverter and control software as part of the system, so the integration interface presented to the vehicle is comparable to a conventional drive.

It adds a degree of freedom and the controls are genuinely more complex than a single-excitation machine. That complexity is contained within the Talos inverter and software stack rather than exposed to the vehicle integrator. It is precisely why we supply the controls rather than shipping a bare machine.

No. Flux management is continuous and handled in software. What is exposed to the user are the different operating profiles, if any.

With a Talos in-house digital position sensor, developed to replace conventional Tamagawa-type resolvers. Against that baseline it delivers lower component cost, lower tooling cost and shorter integration development time.

Dual Flux v1.0 uses neodymium magnets, in substantially reduced quantity: up to 80% less magnet material than a comparable permanent magnet machine. This dramatically reduces exposure to rare earth supply chains but does not remove it.

The v2.0 architecture, defined in 2025, targets ferrite magnets in place of neodymium and is the route toward near-total independence from rare earth supply.

This is programme-specific. CFD and thermal analysis are carried out in-house and specialised cooling validation rigs are used during development. A cooling strategy is defined for the specific application rather than quoted generically.

The architecture scales across power classes and depends highly on client specifications. Built and tested Dual Flux hardware to date centres on a 55 kW class machine, with 15 kW electric and hybrid variants in development for the two-wheel market.

Two patents pending.

Founded in 2018 as an engineering services provider, Talos has built significant experience in electric powertrain and motor development. This accumulated engineering expertise enabled the company to begin developing its proprietary architecture in 2023, following the establishment of the Banbury Technical Centre and initiation of Innovate UK Project Hipermod. The technology subsequently progressed through a motorcycle OEM partnership in 2024, culminating in the definition of the Dual Flux v2.0 architecture in 2025.

Twelve complete vehicle projects, over 10,000 hours of optimisation study, and more than 40 years of combined engineering experience across the founding team, with a background including JLR, Arrival and Volvo. Two patents pending.

To make Dual Flux the de facto architecture for lower-power market segments (motorcycles, ATVs and small cars) by 2030 and the prevailing solution for the high-power vehicle market by 2035.

info@talos-technology.com