Vision system boosts productivity, reduces costs at automobile assembly plant
A leading global automobile manufacturer assembles 120,000 vehicles annually-one to two cars per minute-at a plant in China, and now has increased productivity by six times by using a two-camera high-speed machine vision system.
At the start of the assembly process, large vehicle parts must be placed on the production line. Originally, parts were moved into production by equipment that was operated manually, which was a cumbersome, time-consuming process that held the potential for risk to the operator as well as to the part itself if it wasn’t moved correctly.
"The manufacturer worked with an international provider of automation technologies to automate this important step of the production process, with the goal of increasing productivity and reducing costs. The company designed robotic grippers to lift the vehicle parts onto the assembly line automatically," explained Tony Zhu, general manager of system integrator Check (Shanghai) Automation System Co., Ltd.
"Challenges still persisted, however, because if the robotic gripper didn’t attach to a part in the right place, usually at designated marks, the part could tilt and be suspended from the robotic arm, or be dropped off the line. Identifying the right position at which to grab the part was so critical that the manufacturer considered using vision technology to locate the gripper arms accurately," Tony Zhu said.
Achieving the right resolution
The automobile manufacturer tested several vision systems, but finding a solution that offered the high resolution and image quality needed was difficult. In response, Tony Zhu and the team at Check Automation System Co. proposed a vision system, vision software, and high-resolution 2 MB 1,600 x 1,200 pixel cameras.
The manufacturing facility operates a number of assembly lines. Interference from other lines impacted the quality of the images generated by other vision systems, but the high resolution available with a machine vision system provided clear and accurate images, critical to the success of the vision application, which has now been deployed on two assembly lines in the plant. The vision system and two cameras process images from the two assembly lines simultaneously. Triggered independently, the two cameras take images at a rate of 15 frames per second, capturing multiple positions at the same time, first to identify the type of vehicle part being moved and then to ensure that the four robotic arms will grab the part in the correct locations. The flexibility to accommodate two types of inspections simultaneously is believed to be unique to the vision system used.
The system supports asynchronous acquisition of images with vision software, which uses its pattern-matching tools to analyze the images to identify the coordinates of all positions and ensure alignment. The software also shares the images with the manufacturer’s database, which facilitates communication between the vision system and the facility’s existing technology, streamlining the production process and ensuring quality control.
If the positioning is accurate, the vision system communicates with robotic arms via built-in I/O connections and RS232 industrial standard communications protocol to guide the robotic arms to lift the part onto the assembly line.
Higher speed, less cost
"The speed at which images are taken and processed has helped the manufacturer automate the assembly line, improve productivity, and reduce costs overall. Before implementing vision technology, they were able to grip one vehicle part every five to six minutes. With the robotic gripper equipped with vision guidance, they can grip a part every two minutes, and because they can do so across two independent production lines simultaneously, they’re increasing productivity by about six times overall." continued Tony Zhu.
"The implementation of a vision system has also reduced the potential for error significantly, protecting the safety of operators and eliminating the risk of damage to the vehicle part being processed. The customer reports a high rate of satisfaction with the solution, which is currently in use on two assembly lines," he said.
The manufacturer also identified a number of key advantages of the vision system:
- Unlike other vision software tried, the software used provided unmatched calibration capabilities to minimize the barrel distortion common with such a large object. Without effective calibration, it would be impossible to ensure accurate alignment of the robotic arms and the designated positions on the vehicle part. Calibration capabilities are complemented by robust alignment algorithm parameters, which help ensure that the marks on the part can be seen clearly so that lifting positions are located accurately.
- The vision system ease of use is an important advantage. The graphical user interface (GUI) allows customers to do virtually everything themselves, from operation to maintenance to training. It also ensures that the solution will scale to meet future demands. Simplicity and adaptability were important.
- Teams from Check Automation System Co. and the vision system provider are available to assist, important to customer satisfaction, Tony Zhu said.
The "vision system has performed flawlessly and become critical to the high level of productivity and accuracy the manufacturing plant is now able to achieve," he said, "but even with such an easy-to-use solution, it is important to know that support is available if needed."
– Steve Zhu is Teledyne Dalsa director of sales for Asia; edited by Mark T. Hoske, content manager, CFE Media, Control Engineering, firstname.lastname@example.org.
- As assembly begins, large vehicle parts must be placed on the production line, originally, manually, at risk to operators and parts.
- The high-speed vision system supports asynchronous acquisition of images with vision software, which uses its pattern-matching tools to analyze the images to identify the coordinates of all positions and ensure alignment.
- Two cameras take images at 15 frames per second, identify the type of vehicle part being moved, ensure that four robotic arms will grab the part in the correct locations, and increase productivity by six times.
Are you blind to multi-fold productivity increases because you’re not considering machine vision?