VIES
FMCW radar can only measure velocity unambiguously up to a limit set by how often each target is sampled. Push a fast target past that limit and its velocity folds back, reading at the wrong speed — a real problem at the closing speeds of motorway traffic. VIES is Provizio’s velocity-extension layer: it recovers the true, de-aliased velocity of every detection — including fast oncoming traffic.
Applications
Wherever targets outrun a radar’s velocity limit.
ADAS
Highway closing speeds and fast oncoming traffic routinely exceed a radar's native unambiguous velocity. VIES keeps those targets at their true speed for ACC, AEB and lane-change decisions.
Rail
Trains and trackside intruders move fast relative to the sensor. VIES de-aliases their velocity for level-crossing protection and forward obstacle detection.
Traffic infrastructure
A fixed roadside sensor watches vehicles flash past at full road speed. VIES recovers each vehicle's true velocity for tolling, incident detection and V2X.
Aerospace & UAS
Drones and other aerial targets close at speeds that wrap a radar's Doppler. VIES extends the measurable velocity range for tracking and counter-UAS.
How it works
How VIES recovers a target's true velocity.
STEP 01
List the candidate speeds
A radar can only read speed correctly up to a certain limit. Anything faster wraps around and shows the wrong speed — much like a clock hand that can't tell 1 o'clock from 13. Because that wrap-around is completely predictable, every measurement points to a small, known set of possible true speeds. VIES works out that shortlist for each detection.
STEP 02
Pick the real speed
Only one option on the shortlist fits the rest of the scene. VIES checks each candidate against the target's direction, against the tell-tale pattern stationary objects always leave, and against how far the target moved between frames — then keeps the one that agrees on all three. The result is a single, correct speed for every detection.
STEP 03
Share with downstream perception
It all happens inside the radar's own processor — the same chip that builds the point cloud — so there's no extra hardware or software to bolt on. Everything downstream sees the usual point cloud, except each point now carries the correct speed instead of a wrapped one.
Integration
On the chip, in the DSP chain
VIES runs as a layer in the radar's DSP chain on the SoC. Its output is the same point-cloud format as without VIES — just with corrected, de-aliased velocity per return. Downstream perception code sees no integration change; it simply gets the right velocity to work with.
- Runs on the radar SoC
- Same point-cloud format downstream
- Velocity span is tunable against compute budget
- No perception-layer changes required
Questions, answered
VIES FAQ
Velocity aliasing. Any FMCW radar can only measure velocity unambiguously up to a limit set by how often each target is sampled. A target moving faster than that limit — typically fast oncoming traffic — has its velocity wrap to a wrong value, and because velocity and angle are coupled, its position is thrown off too. VIES recovers the true, de-aliased velocity of every detection.
Hypothesis selection. A wrapped measurement can only correspond to a small, known set of candidate velocities, so VIES forms those candidates for each detection and then keeps the one that is physically consistent — with the target's measured angle, with the way static scenery must line up along a velocity-versus-angle trend, and with how the target moves in range between frames. It is all signal processing in the radar's DSP; no special hardware.
It doesn't reduce range or angular resolution. The real trade-off is compute: testing more candidate velocities costs DSP cycles, so the unambiguous-velocity span is tunable against the processing budget — for example weighting it toward fast oncoming traffic, which matters most for safety.
On the radar SoC. It's a layer in the DSP chain; downstream perception code sees the same point-cloud format, just with corrected velocity.
Unambiguous velocity, on the chip.
If your radar is hitting velocity-ambiguity limits — fast oncoming traffic breaking up, or speeds beyond the native unambiguous range — VIES recovers the true velocity, on the chip.