ALPHA VISION
Live physics simulation · every figure on this page carries its class

Sport is full of numbers.
Very few of them are measured.

Most tracking infers motion from pictures. A camera sees where an object was in two frames and calculates the rest, which is why it needs light, calibration and a clear line of sight. Doppler radar measures velocity itself, directly, with nothing attached to the object. One measurement layer. Four sports on top of it.

Nothing on the object
No light required
Measured, not inferred
Every number here is labelled
Golf
Object in the beam now
0
Radial speed, m/s
156
Position updates per second design
74
Detection range this object, m
Tracking
Measurement state
01 · The principle

Seeing is not measuring

The distinction sounds pedantic until you look at what it costs. An optical system does not measure speed. It measures where something appeared in one frame, where it appeared in the next, and divides. Every error in those two positions is amplified by the division, and the whole method stops working the moment the light does. A Doppler radar measures velocity itself: the frequency of the returned signal is shifted by the motion, and that shift is the measurement. Nothing is inferred, nothing is attached to the object, and darkness and rain are irrelevant to it.

What optics does

Infers velocity by differentiating position across frames. Needs light, calibration and line of sight. Excellent at recognising what a thing is, and at seeing many slow things at once.

What radar does

Measures velocity directly from the Doppler shift, then resolves range from the chirp. One fast object in the dark, in the rain, at distance, with nothing on it.

Where radar stops

It sees motion, not identity, and anything moving slower than about 2 m/s towards or away from the sensor disappears into ground clutter. A putt is invisible to it. We say so on the page where it matters.

Why we can build it

The affiliate this team comes out of has built Doppler instrumentation for ballistic test ranges and space surveillance for decades. The physics is not new to us. The sport is.

02 · The measurement layer

One layer. Change the object, not the physics.

The same sensor and the same signal chain measure a golf ball, a puck, a football and a race car. What changes is how strongly the object reflects, how fast it travels, and therefore how far away it can be seen. Pick an object and watch the measurement run. The panel underneath is not marketing copy: it is what this layer can and cannot deliver for that object, with the class of every claim shown.

MEASUREMENT SCENE · side elevation · true 1:1 scalesimulated physics
measured by the sensor outside the measurement, projected below the clutter gate

03 · Where it goes

Four sports, at four different stages

We are honest about which of these is a product and which is a programme. Golf is the business. The rest are the same layer pointed at a different object, and we say plainly how far each one has come.

Broadcast is a layer, not a sport

Measured numbers are worth more on screen than estimated ones, in any of the four. We treat broadcast as an output of the measurement layer and always sell it with the rights holder, never around them.

What we say no to

Plenty. Sports where the object is slow and curved, where a phone camera already does the job well enough, or where the measurement contract is already locked to somebody else for years. A sensor company that says yes to everything measures nothing well.

04 · Technology

A compact sensor, a grid, and a heritage of measurement

Deliberately simple on the outside: a small sealed unit, several of them where one will not do, and a fusion layer that turns many sensors into one picture. The numbers below are the real ones, each with its class.

The sensor

24 GHz FMCW

A frequency modulated continuous wave radar in the licence free band, in a sealed unit with no moving parts and no lens to clean. A second band is in development for the long range case.

The grid

Many, not one

No single radar covers a golf course or a circuit. Coverage comes from several units and a fusion layer that merges them into one track. Building and synchronising multi radar systems is exactly the heritage this team comes from.

The fusion layer

One picture

Synchronised sensors, multilateration, and a tracker that rejects ghosts and reflections. Identical for every customer, and the reason the layer can serve four very different products.

The heritage

±0.05 % measured

Certified velocity measurement on objects from 30 to 10,000 m/s. This is the affiliate instrumentation heritage the team builds on, not a specification of the sports product.

Detection range

33 / 74 / 132 m design

To a golf ball at 24 GHz: mobile power, fixed power, and fixed power against a higher reflectivity ball. Calculated from the link budget, not yet measured on a range. Larger objects are seen very much further, which is why the constraint differs by sport.

Range resolution

1.5 m design

Set by the chirp bandwidth. This is how finely the sensor separates two objects in depth. It is not a position accuracy figure, and anybody quoting centimetres of range resolution at this bandwidth is quoting something else.

Position, fused

0.25 m class simulated

Median across a full race simulation on real circuit geometry, from several sensors fused. It comes out of our simulation chain and has not yet been measured at a live event.

Launch accuracy targets

±2 m/s · ±1° design

Ball speed and launch angle for the golf case, with carry projected to about ±4.6 m in still air. These are the targets the prototype is being built against and the numbers the first measurement campaign will either confirm or move.

Update rate

50 to 150 Hz design

Per sensor, with several sensors seeing the object at any moment. Independent of light, weather and satellite visibility.

What we do not claim

Honesty

We do not claim centimetre precision, we do not claim validated spin, we hold no CE or FCC certification yet, and we have no measured accuracy figures from a live venue. Officiating decisions rest with governing bodies. Every figure above carries its class, and the campaign that turns the design targets into measured ones is dated below.

05 · Where we are

From design targets to measured ones

The honest position, in public. We are a hardware programme with a working signal chain, a validated physics model, and no measured venue data yet. That last part is what the next twelve months are for.

2026 · done

Signal chain and physics validated

The measurement chain is running on development hardware, and the physics model behind every simulation on this site is built on decades of Doppler instrumentation from the affiliate we come out of.

2026 · in progress

Measurement campaigns

Radar cross section and dynamics measured on real objects with instrumentation grade reference radars, in golf and in motorsport. Until those land, every performance figure on this site is a design target or a simulation, and is labelled as such.

January 2027

Working prototype in public

A prototype measuring the vital parameters, in the enclosure it will ship in, shown at the industry's main annual gathering.

2027

Certification and first installations

CE and FCC, then the first instrumented venues running the full measurement catalogue alongside the systems they already have, scored head to head.

2028

Commercial operation

The same layer, the same fusion, across sports and geographies.

Talk to us

Bring your hardest measurement problem.

A 60 minute technical session: your questions, our physics, and a walk through the chain behind every number on this page, including the ones we have not measured yet.

Request a technical session

Ole Kjærgaard Schrøter · Alpha Vision · oks@weibelventures.com · +45 20 97 56 20