In this post we are going to talk about Bin picking. As you already know, we have talked a lot about the good couple that robotics and computer vision make. Bin picking is one of the applications that we can carry out by joining these two technologies. Next, we tell you what bin picking is, its advantages compared to traditional pick and place applications, and its advantages and applications.
What is bin picking?
Bin picking is an application based mainly on two technologies: robotics and computer vision. The union of these two technologies makes it possible to extract parts from a container that are arranged in a chaotic manner. The pieces are extracted automatically by a robotic arm which receives orders based on the analysis carried out by the computer vision system. The vision system analyzes the scene and recognizes the positions of the pieces. In addition, it determines which part is the most suitable to extract in each iteration, thus optimizing the entire process. The position of the pieces is established in a 3D space, so the robotic arm is able to pick them up in any position.
The bin picking process requires the following steps:
- Recognize and locate the parts within the container. It is necessary to know their position and orientation.
- It must recognize the distance the container is to position and calculate its trajectories in 3D space.
- Optimize the extraction path.
- Avoid obstacles in the environment during their movements.
- Remove the object or part properly without damage.
- Position the object or piece at its destination.
All these tasks are carried out autonomously and in reduced cycle times.
Bin picking also allows you to manipulate pieces or more heterogeneous material that until now was done manually. This allows automating other types of tasks. In addition to the vision system, the robotic part is important since it is necessary to adapt the end effector or gripper to the type of parts to be extracted and the extraction paths must be optimized to optimize the process as much as possible.
In short, Bin picking allows us to reliably recognize and extract parts, regardless of their geometry or material.
mente de su geometría o material.
Bin picking vs traditional pick and place
The traditional pick and place is based on 2D vision instead of using 3D vision, the robot bases its movements on previously fixed trajectories. It is necessary to program the paths that will be repeated by the robot during the operation. The pieces cannot be positioned and oriented randomly since the robot will carry out the operations for which it is programmed, that is, place from the first to the nth piece from one position to another.
To sum up:
- Locate the part with the orientation and position already known.
- Take it and place it in the new location.
In contrast, bin picking is a complex process in which randomly placed parts must be removed from a container, without disturbing the environment. For this, 3D vision is used that allows to collect parts in any position and orientation. The robot will adjust its position and trajectory to extract and place the part in its new position and with the correct orientation.
- Locate the part that is in an unknown position and orientation inside a box / container.
- Autonomously plans the way from the collection to where you have to leave it.
- Enter the box with the appropriate position to take the piece.
- It does not break or hit pieces adjacent to the piece you are picking up.
- It comes out of the box / container and places it in the target position.
The pick and place has been and continues to be very useful to carry out operations in which the part has a certain order. Automating this type of task improves the ergonomics of the production lines since, with them, repetitive movements and the manipulation of dangerous pieces, among others, are avoided. However, bin picking goes further and allows more complex pick and place tasks to be automated, without the need for pre-ordering and with complex part configurations.
Advantages and benefits of bin picking in the industry
A bin picking system has a 3D vision system that allows parts to be located, positioned and oriented, and a robotic arm. Vision robotic systems have many properties that make them adaptable to production environments. Among other things:
- They are very flexible; the vision allows robots to adapt to the environment in which they work without losing efficiency.
- They are very adaptive, they can be rescheduled for other tasks, moved around or even cover a greater volume of work.
- The supervision they need is minimal since the vision system acts as a guide.
- They are very safe because the computer vision and the integrated sensors allow them to coexist with other equipment and with the staff.
- Tasks are always performed with the same precision, even the most repetitive.
- The reduction of security costs, as the system does not need large fences or security equipment. They coexist in work environments with staff and other equipment.
- Increase in speed and precision, parts should not be pre-ordered, so the download or movement is faster and more efficient.
These systems are already being used in many production tasks and have been tested in continuous environments. They also generally demonstrate a high return on investment in production processes that are constant or require extensive training to perform.
Applications in industry
Pick and place applications are present in many different industries. Likewise, bin picking applications can be implemented in all types of sectors such as automotive, logistics or pharmacy, among many others.
As we have told you, the main function of this type of system is to automatically extract pieces that are placed randomly and that can have complex morphologies. For example, pieces with different textures, very complex geometries or made of different materials.
Pickings can be done in all kinds of environments. A pick and place can be made from a belt with randomly placed pieces, or a pick and place with pieces placed in a chaotic container. Even depalletizing tasks can be performed.
One of the projects that we have carried out in ATRIA required the location and extraction of different pieces placed in chaotic in containers of different sizes, in order to carry out a final assembly. The automation of this task would have been unfeasible without the location performed by a bin picking system and allowed to reduce the cycle time and improve the ergonomics of the workstations by reducing repetitive movements.
Do you want to use Bin picking in any of your processes? Contact us!
- Computer vision and robots, the perfect couple?
- Industrial robotics: traditional, collaborative and adaptive
- Computer Vision
- 7 applications of computer vision