Computer vision, particularly through 3D cameras, offers significant benefits in industrial quality control applications, including cost-savings, improved efficiency and streamlined inspection processes, says test and measurement solutions provider TANDM business development engineer Dylan Naidu.

Aligned with increased automation trends, TANDM has added industrial and embedded vision technology developer Basler’s high-resolution Stereo ace 3D industrial camera to its offering.

Combined with AI-powered image processing, the camera improves defect detection, consequently accelerating identification while reducing false positives by distinguishing genuine defects from acceptable product variations.

Noting that quality control still depends on human judgement, Naidu says humans can miss subtle inconsistencies or defects.

Computer vision and 3D cameras fill this gap through precise, consistent and real-time insights, significantly reducing errors across various sectors, including logistics, mining and automotive. This, in turn, significantly reduces risks and expenses.

“Catching defects early is critical, as errors in production can cost millions,” adds Naidu, adding that this can also damage brand reputation and trigger recalls.

In large-scale distribution centres, Basler’s 3D cameras can instantly measure package dimensions, optimise storage allocation and verify barcodes for order accuracy. This reduces manual labour requirements, improves fulfilment speeds and prevents costly shipping errors.

Beyond defect identification and barcode scanning, 3D and 2D cameras enable robotics systems to perceive their surroundings more effectively than manual inspection methods, which are often time-consuming and prone to error, he explains.

By providing sight for robotics systems, 3D cameras also allow for industrial automation tasks such as palletising, object recognition, optical character recognition code scanning and pick-and-place operations.

With the integration of AI software and more traditional image processing, the cameras can recognise colours and perform a variety of other inspection tasks consistently and repeatedly.

“These solutions were specifically de- signed for warehousing, where time efficiency matters most,” he says.

The high-resolution imaging of the Basler Stereo ace 3D industrial camera is factory pre-calibrated, ensuring “seamless integration” and flexible interface abilities.

These capabilities are supported by an advanced software suite that offers plug-and-play functionality through Basler’s Pylon interface, which is also used for Basler’s 2D cameras. This allows for easy setup and deployment across robotic platforms.

With AI and traditional processing, users first define what a ‘good’ product looks like, accounting for acceptable variations. Once that definition is established and the system’s models are trained, the system can automatically identify deviations and manage the inspection process based on those parameters, explains Naidu.

Once an application’s parameters are defined, the cameras deliver identical outputs based on identical inputs every time, offering a level of repeatability “far superior” to manual inspection.

Basler also provides an open software development kit, allowing for integration into third-party software applications. Through additional development input from software engineers, this capability enables users to embed camera data streams into broader workflows or management systems.

Naidu anticipates rapid technological evolution, as cameras are expected to become significantly smarter within the next five years.

“If I consider how the software has evolved over the last five years, and how the hardware has developed, I foresee this technology will become even more widely adopted.”

As adoption continues to grow, increased investments in automation technology will allow for software improvements, alongside increased affordability, offering significant value for industries that leverage such technology, Naidu concludes.