Vision-guided robotics uses 2D or 3D cameras to tell a robot exactly where each part is and how it is oriented, so the robot can pick parts that are not precisely positioned: off a moving conveyor, out of a tray, or jumbled in a bin. It removes the need for expensive fixtures and part feeders.
A robot on its own repeats taught positions precisely, but it is blind: the part must be exactly where it expects. Vision-guided robotics adds a camera that finds each part and sends its position and orientation to the robot, so the robot adapts its path every cycle.
2D vision is enough when parts lie flat on a conveyor or tray. 3D vision measures depth as well, which is needed to pick randomly stacked parts from a bin, known as bin picking, or to unload mixed cartons from a pallet.
A camera above the conveyor, tray or bin captures the scene and finds each part, in 2D or in 3D.
Software calculates where the part is and how it is turned, and chooses a part the gripper can reach.
Hand-eye calibration converts camera coordinates into robot coordinates; an encoder adds conveyor movement.
The robot plans a collision-free path, picks the part and places it in the target position.
The camera can confirm the pick succeeded and the part was placed correctly before the next cycle.
Pattern matching finds flat parts on a conveyor or tray and gives their X, Y position and rotation to the robot.
Structured light, stereo cameras or time-of-flight sensors produce a point cloud so the robot can pick parts stacked or jumbled in a bin.
A calibration routine precisely relates the camera to the robot, whether the camera is fixed above the work area or mounted on the robot arm.
An encoder on the conveyor lets the robot pick moving parts by predicting where each part will be when the gripper arrives.
Two example project scopes showing how the station, the checks and the outputs are defined. Every plant is different, so the final configuration is confirmed after a feasibility study on your own parts.
The situation: A food producer packs products into trays by hand at the end of a production line.
The situation: A component maker loads machined blanks into a machine by hand from a bulk bin.
Industrial cameras, lenses, line-scan cameras and code readers are available through our machine vision catalog, quoted with delivery to Sri Lanka. We select the hardware for your application during the feasibility study.
Browse the hardware catalogEvery product, surface and line behaves differently under a camera, so every project starts with evidence from your own material.
A site visit to understand your product, line speed, quality criteria and where the check should sit.
We image your good and defective samples and show what the system can reliably detect. This is the go or no-go.
One station on one line, running beside your inspectors until results match and your team trusts them.
More lines, integration with your PLC and systems, and local support from our Sri Lankan engineering team.
Vision error-proofing (poka-yoke) for assembly lines: confirm every component is present, correct, oriented and seated, count parts, and keep an image record of each assembly for traceability.
Optical sorting and grading for tea, spices, coconut, nuts, grains and fresh produce. Colour, NIR and hyperspectral vision remove stalks, foreign matter and defective product, and grade by colour and size.
Sort fabric rolls, cut panels and washed garments into buyer shade bands with CIELAB measurement and 555 codes.
Every machine vision application we design and build, in one place.
Bin picking is a robot picking parts that are randomly piled in a bin. It needs 3D vision to measure where each part is and how it lies, and software that chooses which part the gripper can reach without collisions.
Use 2D when parts lie flat in a known plane, such as on a conveyor or tray. Use 3D when parts are stacked, jumbled or vary in height, such as in a bin or on a mixed pallet.
Usually, yes. Most industrial and collaborative robots accept position data from a vision system. We check the robot model and controller during the feasibility study.
It is the procedure that relates the camera to the robot so that a position found in the image converts into the exact position the robot should move to. It is done when the cell is commissioned and checked whenever the camera or robot is moved.
Yes. With conveyor tracking, an encoder measures belt movement so the robot knows where each located part will be when the gripper reaches it.
Tell us what you need to check and share some good and faulty samples. We will show you what a camera can reliably see before you commit to anything.
Photographs show representative plants and equipment, not Cerox installations or client sites. Diagrams by Cerox Engineering. Photo credits, via Wikimedia Commons: Robot picking from a crate by Mixabest, CC BY-SA 3.0; Robotic packing cell by Tecnowey, CC BY-SA 3.0.