Metal Parts in Bins
Locate machined, cast, stamped or sawn components and calculate suitable picking points.
3D Stereo Vision Camera for Robotic Bin Picking
In many production environments, parts arrive randomly in bins, crates or pallets instead of being neatly positioned for automation. For a robot, this makes picking difficult because it must understand the part’s position, orientation and accessibility before it can move safely.
A 3D camera for robotic bin picking gives the robot the depth data it needs for object localisation, pick point calculation and robot guidance. Stereo vision cameras generate point clouds that help automate bin picking, machine tending and depalletising applications with greater flexibility than fixed mechanical feeding systems.
What Problem Does Robotic Bin Picking Solve?
Many automated production lines still rely on manual handling or mechanical feeding systems. These solutions can work well when the part range is limited and every object can be presented in a predictable position. However, mechanical feeding can become complicated when:
- parts vary in size or geometry
- objects arrive randomly in bins or crates batches change frequently
- parts are difficult to orient mechanically
- manual loading is repetitive or ergonomically demanding
- the process requires more flexibility than a fixed feeder can provide
Robotic bin picking solves this by allowing a robot to pick parts directly from an unstructured or semi-structured container. Instead of requiring every part to be pre-positioned, the robot uses vision data to decide which part can be picked next.
The main challenge is not simply detecting the object. The robot needs usable 3D information. It must know the object’s position, orientation and accessibility before it can perform a reliable pick. This is where 3D stereo vision becomes valuable.
Why 2D Vision Is Not Enough for Most Bin Picking Applications
A 2D camera (also called area scan camera) can be useful for inspection, classification, code reading or part detection. But in robotic bin picking, the robot must move through real physical space. It needs in-depth information.
A 2D image can tell the system what the camera sees. A 3D camera helps the system
understand where the object is in space. For robot picking, depth data is used to determine object height, object position and orientation, visible surfaces, possible gripping points, distance to bin walls, part overlap, collision risks
Without 3D data, it is difficult to calculate a reliable robot path or determine which part is actually accessible. This is why 3D machine vision is commonly used for bin picking, machine tending, depalletising and other robot-guidance applications.
Where Robotic Bin Picking Is Used
Robotic bin picking is used when parts arrive randomly positioned in bins, crates, containers or small load carriers. Instead of using a mechanical feeder to present every part in the same orientation, a robot uses 3D vision to locate the next accessible part and pick it from the container.
This makes bin picking especially useful where manual loading is repetitive, part orientation changes between cycles or flexible automation is needed for different part types.
Plastic Parts in Crates
Transfer injection-moulded parts to conveyors, fixtures, inspection or packaging processes.
Machine Tending
Pick raw parts from small load carriers and load CNC machines, presses or fixtures.
Industrial Components
Handle brackets, housings, connectors, fasteners and other randomly positioned components.
Large Crates and Containers
Cover larger picking areas using a suitable field of view and working distance.
How Stereo Vision Cameras Supports Robot Picking
A typical 3D stereo vision robot-picking workflow looks like this:
-
1
The 3D camera captures the bin, crate or pallet.
-
2
The stereo vision system generates depth data.
-
3
A 3D point cloud represents the visible object surfaces.
-
4
Software identifies parts, surfaces or grasp candidates.
-
5
The system calculates pick points and object pose.
-
6
Collision zones and gripper access are checked.
-
7
Robot coordinates are sent to the robot controller.
-
8
The robot performs the pick and places the part in the next process step.
The practical output is the point cloud. This 3D data gives the software information about the visible surfaces in the scene. From this, the system can determine which object is accessible, where the object is located and how the robot should approach it.
For many applications, the 3D camera is mounted above or near the bin. In other cases, the camera may be mounted on the robot arm. The best mounting position depends on the bin size, working distance, robot movement, gripper design and required field of view.
Why Point Cloud Quality Is Critical for Reliable Bin Picking
Robotic bin picking is only as reliable as the data behind the robot movement. If the point cloud is incomplete, noisy or poorly matched to the application, the robot may fail to identify a valid pick point.
Several factors influence 3D data quality:
- Surface texture
- Reflections
- Object colour
- Working distance
- Field of view
- Camera angle
- Part overlap
- Bin fill level
- Ambient light
- Required cycle time
- Gripper accessibility
For example, low-texture or reflective parts can be difficult for stereo matching because the system needs visible differences between both camera views. Dark plastic, shiny metal and repeated object features can also make 3D capture more challenging.
This is why the application should be evaluated before selecting the camera. The right 3D camera for robotic bin picking depends on the real parts, real bin, real working distance and real cycle time.
VA Imaging 3D Stereo Vision Camera Powered by Ensenso
VA Imaging 3D stereo vision cameras powered by Ensenso combine advanced stereo vision technology with camera configurations selected for demanding industrial automation and robotics applications. The cameras generate detailed depth data and reliable point clouds for tasks such as robotic bin picking, machine tending, depalletising and 3D object localisation. Depending on the application, VA Imaging offers standard VA Choice models as well as built-to-spec configurations tailored to the required working distance, field of view and 3D data quality.
Key Selection Factors for a Robotic Bin-Picking Camera
Choosing the right 3D camera for robotic bin picking is not only about selecting a camera model. It is about matching the full vision setup to the application.
Part Size and Shape
Small, complex or overlapping parts usually require more detailed 3D data than large, simple components. The shape of the part also affects pose estimation and the identification of a stable gripping point.
Surface Finish
Reflective, dark, transparent or low-texture surfaces can be challenging. Surface finish influences camera selection, acquisition settings, mounting angle and the need for projected texture.
Bin Size and Fill Level
The camera must cover the required bin area while maintaining enough detail for localisation. Larger bins may require a wider field of view, while smaller parts may need higher point density.
Working Distance
Working distance affects field of view, point density and depth performance. The selected camera should match the actual mounting position in the robot cell.
Field of View
The field of view should cover the required picking area, not necessarily the entire environment. The correct FOV is usually the area in which valid picks can be made.
Robot and Gripper Design
Gripper size, suction cup layout, jaw opening, approach direction and clearance all influence what counts as a valid and accessible pick.
Hand-Eye Calibration
Hand-eye calibration connects the camera coordinate system to the robot coordinate system, allowing the robot to move accurately using the calculated 3D position.
Software Integration
The complete system also needs software for point cloud processing, object localisation, pose estimation, grasp planning and communication with the robot controller.
How to Choose a 3D Stereo Vision Camera for Robotic Bin Picking
For robotic bin picking, the camera must capture the bin clearly enough for the system to identify visible parts, understand their position and calculate usable pick points. The right stereo vision camera depends on the bin size, part dimensions, surface finish, mounting position and the accuracy for reliable picking. Working distance and field of view help define whether the camera can see the required picking area, but the final choice also depends on the bin-picking conditions. A compact bin with small load carriers may need a different camera approach than a larger container, mixed parts or a depalletising task.
B57 3D Camera VA Choice for Compact Robotic Bin Picking
The B57 3D Camera VA Choice is a strong starting point for compact robotic bin picking, small load carrier picking and machine tending from bins.
This camera option is especially relevant when space in the robot cell is limited and the camera needs to be mounted close to the picking area. It is a practical choice for smaller bins and crates that require reliable 3D data from a compact camera setup.
5MP
Stereo vision
GigE
Camera interface
FlexView
Pattern projection
IP65/67
Industrial housing
Typical applications
Particularly relevant when the camera needs to operate close to the picking area or fit inside a compact automation cell.
- Compact robotic bin picking
- Small load carrier picking
- Part picking from crates
- Robot guidance in compact cells
- On-arm vision and eye-in-hand applications
VA Choice configurations
Choose the required picking area
Compare the two predefined B57 configurations by working distance and field of view.
B57 3D Camera VA Choice
Wide-area compact configuration
B57-6/900-N/3-BL
Working distance
850 mm
Field of view
800 × 600 mm
This configuration provides wider scene coverage for compact robot cells where the camera needs to capture a larger bin or picking area from a relatively short mounting distance.
View configuration
B57 3D Camera VA Choice
Focused compact configuration
B57-12/975-N/3-BL
Working distance
900 mm
Field of view
400 × 300 mm
This configuration provides a more focused view for smaller picking areas where detailed 3D data is required within a clearly defined working zone.
View configuration
Selection advice:
The final configuration should match the part dimensions, bin size,
mounting position, surface characteristics and required point-cloud detail,
rather than being selected only by field of view.
C67 3D Camera VA Choice for Colour-Enhanced 3D Bin Picking and Larger Containers
The C67 3D Camera VA Choice is suitable for robotic bin picking applications that require detailed 3D data, larger picking areas and additional RGB colour information.
With 5MP stereo vision, an integrated 8MP RGB colour sensor and 5GigE connectivity, the C67 supports demanding robot guidance, object localisation, depalletising and automated handling applications.
5MP
Stereo vision
8MP RGB
Colour imaging
5GigE
Camera interface
IP65/67
Industrial housing
Typical applications
Particularly relevant for applications that require wider coverage, colour-supported 3D data or longer camera working distances.
- Robotic bin picking
- Large-bin and crate picking
- Depalletising and pallet handling
- 3D object localisation
- Colour-supported robot guidance
- Industrial 3D measurement
VA Choice configurations
Choose the required picking area
Compare the four predefined C67 configurations by working distance and field of view.
C67 3D Camera VA Choice
Focused compact configuration
C67-12/28/975-S/7
Working distance
900 mm
Field of view
400 × 300 mm
A focused configuration for smaller picking areas that require detailed 3D data and colour information within a defined working zone.
View configuration
C67 3D Camera VA Choice
Balanced compact configuration
C67-8/28/975-S/7
Working distance
900 mm
Field of view
600 × 400 mm
A balanced option for medium-sized bins and picking areas where scene coverage and detailed 3D acquisition are both important.
View configuration
C67 3D Camera VA Choice
Wide-area compact configuration
C67-6/16/975-S/7
Working distance
900 mm
Field of view
800 × 600 mm
Provides wider scene coverage for larger bins, crates and automation cells while maintaining a relatively short working distance.
View configuration
C67 3D Camera VA Choice
Large-volume configuration
C67-8/16/2200-M/7
Working distance
1800 mm
Field of view
1200 × 1000 mm
Designed for larger containers, depalletising and wide picking areas where the camera must capture a larger working volume from further away.
View configuration
Selection advice:
The final configuration should match the bin dimensions, part size,
mounting position, required working distance, surface characteristics,
colour requirements and required point-cloud detail.
Need help selecting a configuration?
Discuss your robotic bin picking application
VA Imaging can help compare the required field of view, working distance, part characteristics, mounting position and 3D data requirements for your application.
Choosing Between B57 and C67 Powered by Ensenso 3D Cameras
When comparing the stereo vision cameras B57 VA Choice and C67 VA Choice powered by Ensenso, start by checking whether the working distance and field of view match the bin size and available mounting position. After that, compare the application requirements.
The camera interface is another important difference between the two series. The B Series uses a standard 1GigE interface, while the C Series supports 5GigE for higher-bandwidth data transfer. This should be considered alongside the working distance, field of view, colour requirements and required acquisition performance.
As a general guideline, the Ensenso B57 VA Choice is a strong starting point for compact, close-range robotic bin picking and machine tending. The Ensenso C67 VA Choice should be considered for larger setups, colour-supported recognition, longer working distance options or depalletising applications.
In many cases, the best camera choice can only be confirmed after reviewing the real part, bin dimensions, surface finish, mounting position and required point cloud quality.
Built-to-Spec Stereo Vision Cameras for Bin Picking
Not every robotic bin picking application fits a standard VA Choice configuration. Some projects require a different working distance, field of view, camera setup or housing approach. In these cases, a built-to-spec stereo vision camera configuration may be more suitable.
Built-to-spec options are especially useful when the bin size, part dimensions, mounting position or required point cloud quality do not match a standard camera configuration.
- Built-to-Spec Ensenso B-Series
A built-to-spec B-Series camera can be considered for compact bin picking applications where the camera needs to fit into a space-limited robot cell or work close to the picking area.
- Built-to-Spec Ensenso C-Series
A built-to-spec C-Series camera can be considered when the application requires colour-supported 3D data, a specific working distance, a larger field of view or a more demanding imaging setup.
Other Ensenso Series for Specific Bin Picking Setups
Depending on the application, other Ensenso series may also be considered.
- Ensenso X-Series (X36 and X30) may be relevant for large containers, large pallets, long working distances, high-volume objects or custom stereo geometry.
- Ensenso N-Series (N46 and N36) may be relevant for compact industrial 3D
imaging, stationary or robot-mounted setups, smaller automation cells or
cost-sensitive 3D vision tasks. - Ensenso CR-Series may be relevant when the application benefits from onboard depth processing, reduced data transfer or
high-performance 3D results directly from the camera.
The best camera option depends on the real bin picking conditions: part size, surface
finish, bin geometry, mounting position, required field of view and required 3D
data quality.
Application Checklist Before Selecting a Stereo Vision Camera
Before selecting a stereo vision camera for robotic bin picking, it helps to define the main application conditions. The most important questions are:
- What are the part dimensions, weight and material?
- Is the surface shiny, dark, transparent or low-texture?
- What are the bin, crate or container dimensions?
- What working distance and field of view are required?
- Where can the camera be mounted?
- What level of 3D detail is needed for reliable picking?
- What cycle time is expected?
- Is colour information useful for the application?
- Are there environmental factors such as dust, vibration or ambient light?
FAQ ABOUT BIN PICKING
Need help choosing stereo vision camera for bin picking?
Share your part size, bin dimensions, working distance, field of view and cycle time requirements with VA Imaging. We will help you select the right stereo vision camera powered by Ensenso and define a practical robot-guidance setup for your application.