An Intro to Vision-Based Binpicking and 2 Industry-Leading Vision Solutions

Let‘s talk binpicking. In this quick article we will get an overview of binpicking itself, its various real world applications, and some solutions you should check out for your own binpicking needs.

But first for those not in the know: What is binpicking? Binpicking refers to the process of automatically identifying and picking up objects from a bin or storage area using robots. It is commonly used in manufacturing and logistics for tasks such as sorting and organizing objects, assembling products, and packaging and labeling items for shipping.

There are a variety of solutions available for binpicking tasks, with one of the most exciting and useful being machine vision. There are a large variety of machine vision solutions and each of these approaches has its own advantages and limitations, and the best solution will depend on the specific needs of the application.

In this article, we'll look at two super solutions for binpicking. We will also explore common problems and issues that can arise when implementing binpicking systems, and provide tips for overcoming these challenges.

Whether you are just starting to explore the possibilities of binpicking or are looking to optimize an existing system, we hope that this article will provide valuable insights and information. If you want to jump right into the world of machine vision, check out Unchained Robotic's selection of machine vision tools here.

Why Binpicking is Hard

Binpicking can be a challenging task for several reasons. But the biggest challenge is simply that the objects being picked up may be oriented in a variety of unpredictable ways, making it difficult for the robot to accurately identify and grasp them. Objects in a bin are usually randomly and haphazardly piled upon one another making it impossible to predict what will be where and in what orientation. When the robot attempts to grab one of these objects it may dislodge others or even end up grabbing more than one.

Further, objects may be stacked or piled densely together, making it difficult for the robot to access them. As you can see, there is nothing simple about binpicking!

Another challenge is that the objects being picked up may vary in shape, size, and material, which can make it difficult for the robot to recognize and accurately grasp them. Though this is relatively uncommon, the majority of binpicking use cases in the real world involve a single identical items in the container.

Another challenge is that the environment in which the binpicking task is taking place may have poor or varying lighting conditions, which can make it difficult for the camera and the machine vision software to identify the pieces. This can naturally lead to errors in the binpicking process.

Lastly, even if the robot manages to do all the above it needs to be able to pick pieces individually. This is especially difficult with workpieces that have complicated geometries because they tend to stick together.

Binpicking is everywhere. Binpicking is unavoidable.

Binpicking is a common task in manufacturing and logistics, and can be found in almost all kinds of production. In manufacturing, binpicking is often used to pick up and move parts or components to a production line, where they can be assembled into a final product. This allows for efficient and automated production of goods, which can greatly improve efficiency and reduce the need for human labor.

In logistics, binpicking is often used to pick up and transport packages or other items for shipping. This can help to improve the speed and accuracy of the shipping process, and can also reduce the need for manual handling of packages.

Binpicking can also be found in other industries and applications, such as agriculture, where it can be used to sort and organize produce or other goods. In healthcare, binpicking can be used to pick up and transport medical supplies, pills, and more. The applications for binpicking are truly endless, hitherto the biggest hinderance has been the technology. But the technology has improved substantially in the last few years.

2 Super Solutions for Binpicking 

Let’s take a quick look at 2 incredibly capable binpicking solutions that you can consider for your own production needs.

Cambrian  - the Powerful Biofocal Solution

This is an advanced machine vision solution that is designed to work with an enormous variety of objects, including those that are dark, see-through, or oddly shaped, and functions even in low-light conditions.

One of the key upsides of the Cambrian Vision solution is its ability to handle objects that may be challenging for other binpicking solutions. This is made possible through the use of advanced imaging and machine learning algorithms, which allow the system to accurately identify and pick up objects even in challenging environments (low light, changing light, etc). The solution can also handle workpieces that are black, dark, semi-transparent, etc. which would be showstoppers for most machine vision solutions.

However, it is important to note that the Cambrian Vision solution does require a CAD file for each object that it will be handling, and there must be significant preparation in advance to ensure that the system is set up correctly. Additionally, the objects being picked up can only vary slightly from the CAD file, so the system may not be suitable for handling objects with significant variations in shape or size.

To function, the Cambrian Vision solution requires two cameras, which are typically mounted on the robot arm. Seeing this solution in person can be very impressive, as it demonstrates truly next-level binpicking capabilities.

Mech-Mind 

The Mech-Mind Mech-Eye system is a 3D vision system available in a range of variants. The line offers a large array of working distances and fields of view and is software-compatible with most major robot brands.

These cameras are designed to be able to identify a wide range of objects, including those that are small, complex, or have a high level of detail.

Mech-Mind Mech-Eye units include two built-in cameras and onboard processing. They are available with specialized software to analyze the objects in the bin and identify their locations and orientations. The robot can then use this information to plan a path by which it can best pick the object.

One of the key distinguishing features of the Mech-Mind Mech-Eye Camera system is its software. This software is compatible with the software of an impressive array of robot manufacturers, making it significantly easier to deply the solution.

Overall, the Mech-Mind Mech-Eye Camera system is a highly effective binpicking solution that can improve efficiency and accuracy in a variety of applications. It is particularly well-suited to workplaces where the objects to be picked might change with relative frequency.

Conclusion

In this article, we have provided an overview of binpicking and its various applications in manufacturing and logistics. We have discussed the different solutions available for binpicking tasks, including vision-based approaches, and the pros and cons of each. We have also explored common problems and issues that can arise when implementing binpicking systems, and provided tips for overcoming these challenges.

For those ready to start their binpicking journey, solutions like the Cambrian Vision solution and the Mech-Mind Mech-Eye Camera system are a great place to start. Each of these solutions has its own unique features and capabilities, and can be a good choice for different applications depending on the specific needs of the task.