Preparing the published article.
Zivid: 3D Robot Vision, R-Series Cameras and Motion Software
Oslo-based robot perception supplier with Zivid 2+ R-series and Zivid 3 cameras, capture workflow considerations and motion software scope.
Company and role in a robot cell
Zivid AS is based in Oslo, Norway. Its current company timeline dates the SINTEF spin-out to May 2015. Zivid supplies perception and motion technology to integrators and robotics companies rather than complete robotic systems. This profile belongs in robot components and control: a camera can support an automation solution without supplying the arm, gripper, process tooling or commissioned cell.
Current portfolio and product generations
The official timeline records Zivid 2+ R-series in December 2024, Zivid 3 in November 2025 and Zivid Motion in March 2026. The current site lists camera hardware, an SDK, Zivid Studio and motion software. Buyers should identify the actual generation and software package in a quotation; an older camera reference does not establish the functions of a newer product or the inclusion of every software option.
Zivid 2+ R-series selection
The R-series combines colour images and 3D point clouds. Zivid positions MR60 for assembly, robot guidance and inspection; MR130 for piece picking, bin picking and parcel handling; and LR110 for piece picking, depalletizing and larger bins. These are different fields of view and intended tasks. Assess the complete working volume, object surfaces and mounting geometry before selecting a model.
Zivid 3 for larger working volumes
Zivid 3 XL250 is presented for stationary mounting and larger warehouse or factory scenes, including depalletizing and bin picking. Its product page describes an 8-megapixel imaging system and long-distance operation around 4.5 metres. Those camera characteristics do not define robot reach, payload or completed picks per hour. The performance of a complete station also depends on localisation, grasp selection and robot movement.
Capture workflow affects throughput
The Zivid knowledge base distinguishes independent 2D capture from combined 2D/3D capture. It explains that ordering, sampling, aperture changes and processing choices affect latency. When 2D and 3D resolutions differ, pixel correspondence must be handled deliberately. A supplier camera capture-time headline should therefore be kept separate from the measured cycle time of segmentation, planning, picking and placement.
Integration and software scope
Ask the integrator to document camera mounting and calibration, computer and network requirements, SDK compatibility, scene processing and the robot-control interface. The current portfolio includes Zivid Motion, but this profile does not assume that a camera purchase automatically includes a motion-software licence or support for every controller. Confirm the licensed functions, supported hardware and responsibility for maintaining the deployed software.
A useful application trial
An evaluation should use representative parts, including awkward surfaces, occlusions and realistic bin fill levels. Record successful localisation and picking separately, with retry rate, damage checks and total cycle timing. Repeat under the expected lighting and temperature conditions. This is a proposed acceptance approach, not a claim that every material will work. Request a dated configuration and support proposal before comparing solution costs.
Official sources
Source review: 11 September 2026. Product descriptions are manufacturer statements; configuration and project outcomes require separate confirmation.

