Longer sight gives a machine more time
Innoviz Technologies launched the extreme-range variant in April 2026. A one-kilometre maximum does not mean the same image for every surface and weather condition; reflectivity, resolution, frame rate and noise determine practical value.
At 130 km/h a vehicle covers about 36 metres each second, while trains and specialised vehicles may need far longer to stop. Early detection helps only if an object is accurately located, classified and tracked rather than appearing as a vague distant return.
Lidar measures the return time of laser pulses
Lidar creates a three-dimensional point cloud by timing reflected laser pulses. Cameras add colour and texture, radar measures speed and handles adverse weather, and lidar supplies geometry, so advanced autonomy often combines sensors.
InnovizTwo uses a solid-state architecture designed for automotive size, cost and mass production. The Ultra variant drives the same foundation differently. Field of view, point density and power at maximum range matter more to integration and calibration than a record number alone.
Rail and heavy machinery differ from passenger cars
A train follows a fixed path and cannot evade an obstacle, so earlier detection of people, animals, debris or infrastructure damage can aid warning and braking. Mines, ports and construction add dust, variable terrain and poor visibility where range can expand the safety envelope.
These smaller markets may pay more for a specialised sensor, but certification, servicing and liability vary. Rail demands long-term availability; defence may emphasise ruggedness and small targets. Success depends on software, interfaces and sector integrators rather than one identically configured box.
The specification needs independent verification
Testing should report detection probability at different reflectivities, range accuracy and resistance to sun, rain, fog and other lidars, including false alarms. A sensor that often sees nonexistent obstacles can trigger braking and lose operator trust despite an ideal-condition record.
“Seeing” must be defined: a few returned points differ from reliably identifying a person or vehicle and predicting its path. Customers need raw measures and performance of the complete detection chain, followed by partner data and series-unit quality.
Price pressure in lidar continues
After high expectations and consolidation, carmakers demand lower cost, smaller size and suppliers able to last through model cycles. Production nominations may take years to produce revenue. Ultra broadens the market while adding development and sales costs.
Innoviz must balance ambition with financial discipline. Customers assess supplier stability, manufacturing and warranty support. Paid non-car programmes could reduce dependence on carmaker schedules; a technical demonstration alone will not.
An opportunity for integration and testing
Czech companies and researchers can contribute sensor integration, simulation, rail diagnostics, robotics and test scenarios. Value comes from combining lidar with cameras, radar, positioning and safe machine control, using sector-specific datasets and verification.
InnovizTwo Ultra shows a move from one automotive product towards a broader perception platform. The memorable kilometre figure must become high-quality integration, service and measurable risk reduction before it delivers a lasting competitive advantage.
Sources and editorial note
The Jews.cz editorial team prepared this article from the public materials below, distinguishing company claims, independently documented facts and editorial interpretation.
