Novel Machine Vision System Ensures Proper Automotive Seat Mold Assembly
Industrial imaging, robotic, and deep learning technologies combine to efficiently inspect highly variable automotive parts.
Automation technologies such as robots and machine vision factor significantly into automotive manufacturing, taking on tasks from painting and welding to material handling and assembly.
An Association for Advancing Automation (A3) Certified Systems Integrator, LEONI Engineering Products & Services (LEPS) utilizes its in-house machine vision development lab and expert team, to regularly solves difficult inspection, assembly, and verification challenges in automotive manufacturing — and beyond. Tasks include wheel and tire validation, robotic guidance, optical character recognition (OCR), and full-scale traceability of manufacturing and assembly processes and even verifying seat foam insert presence and position.
Machine Vision System Design and Services
The company begins with an in-depth analysis and an engineering feasibility study to build a foundation for the vision solution design. Then, using the best machine vision hardware and software for the application, LEPS designs the system, documents it, and trains operators on its use.
Seat mold inspection is a particularly challenging machine vision task. Working within a miniscule profit margin, manufacturers rely heavily on automated inspection systems to verify proper components have been encapsulated into each foam piece. But the increasingly complex design and large variety of seat molds requires an inspection system capable of adapting on the fly.
Seat foam insert assembly involves a lot of components. Manually placed by operators, these can include clips, Velcro, stiffening wires, cloth, and RFID tags placed inside the seat mold cavity before a robot pours polyurethane foam into the mold. A production line may also handle 100’s of part types, which must all be inspected. Variation in mold appearance adds another layer of complexity. For example, molds appear extremely reflective after cleaning and become dark when waxes and dyes are added into the cavity.
Once foam has been poured, most incorrectly placed components cannot be reworked or even detected, so catching defects ahead of time is a top priority for manufacturers, who of course do not want to create waste or damage relationships with the customer. Several automotive manufacturing companies even commit to having an inspection system in place as part of the contract to supply materials for the customer.
Novel Design Enables Precise Inspection
To improve seat mold assembly, LEPS designed a flexible automated inspection system that accommodates several different part numbers, models, and inspection tasks. The system is currently deployed in several different locations.
The process begins right after the operators finish placing components into the cavity. One of the inspection system’s cameras captures images of predetermined points in the cavity where most variation occurs. The system uses a dedicated camera to measure the intensity of the mold so that the main inspection cameras can be offset to create maximum contrast between the components and the mold itself.
The system compensates for different seat geometries and mold depths by strobing an extraordinary amount of light and using a novel optics and lighting setup. This increases depth of field, enabling the system to covers the full range of depth required to inspect all seats and molds, explained Jim Reed, Vision Product Manager, LEPS.
“Whether the cavity is four, eight, or ten inches deep, we can keep that whole range in focus,” he said.
Next, the main inspection cameras verify that correct components are placed in the proper locations in the seat cavity. Once the system software confirms the presence of the components, it issues a pass or fail determination. When a mold passes, the system sends a signal to the programmable logic controller (PLC). The mold then travels to the robot pour station, where it’s filled with foam, the lids close, and the feed is completed. If the system detects missing components, it alerts the PLC to not pour that seat mold. The seat then travels back to the operators, who double-check and correct the component placement. Afterward, the seat is resent through the vision inspection station.
Deep Learning Eases Part Variance Concerns
Part variance can become overwhelming in some instances. For example, molds may have handwriting or machining marks and look quite different from one another even if they are producing the same part. Previously, each mold had to be programmed individually, but LEPS has begun incorporating deep learning into the process. Operators can now program a single part number rather than every seat cavity in house, saving a lot of time in the process.
“These systems are incredibly complicated from a vision standpoint because of the sheer volume and variation between part numbers and models, as well as the programming and changes,” Reed said. “One mold may have 30 inspections being performed on it, so if you take those inspections and multiply it by 200 part styles for 1,500 different molds, you can quickly see how this becomes quite complex.”
By applying our knowledge of “deep learning”, our vision systems can accommodate a lot more variation which reduces the time spent programming and maintaining the system. These and many other advanced vision techniques provide a wide range of solutions for many potentially challenging applications, beyond just the mold insert verification and inspection.
Information on LEPS’ machine vision service can be found here. For questions about LEONI’s machine vision systems and services, contact Jim Reed by email at email@example.com or by phone at 248-766-6844.