Kia
Assembly Line
Hyundai’s Wearable Robotic Shoulder Coming To A Mechanic Near You
Developed by Hyundai and Kia’s Robotics LAB, the X-ble Shoulder isn’t just another gimmicky piece of tech. It’s designed to actively assist workers who have their arms raised for extended periods, reducing shoulder load by up to 60%. For the anatomy enthusiasts, it also reduces anterior and lateral deltoid muscle activation by up to 30%. In other words, it lightens the load both literally and figuratively.
What’s really impressive here is the muscle compensation module that drives the X-ble Shoulder. This clever piece of tech can perform an astounding 700,000 folding and unfolding actions per year, so it’s built for endurance, not just a one-off task.
Obviously, no one’s going to want to work in a hulking exoskeleton all day, so Hyundai has ensured that this device won’t weigh you down. The X-ble Shoulder weighs in at just 1.9 kg (4.1 lbs), thanks to its carbon composite construction. It’s lightweight, adjustable, and designed to fit snugly without feeling like a straitjacket. It comes in two variants: the basic version offers up to 2.9 kgf of assistive force and is best for tasks where posture isn’t fixed, while the adjustable version delivers 3.7 kgf for those needing a bit more muscle.
And Hyundai’s not stopping at the shoulder. The Korean company is also working on an X-ble Waist to assist with lifting heavy loads and reduce back injuries, as well as an X-ble MEX, designed for the rehabilitation of the walking impaired. So, it looks like the future of wearable robots could be a lot more comprehensive than just helping you lift that engine block.
Robust unsupervised-learning based crack detection for stamped metal products
Crack detection plays an important role in the industrial inspection of stamped metal products. While supervised learning methods are commonly used in the quality assessment process, they often require a substantial amount of labeled data, which can be challenging to obtain in a well-tuned production line. Unsupervised learning has demonstrated exceptional performance in anomaly detection. This study proposes an unsupervised algorithm for crack detection on stamped metal surfaces, capable of classification and segmentation without the need for crack images during training. The approach leverages the Vector Quantized-Variational Autoencoder 2 (VQ-VAE2) based model to reconstruct input images, while retaining crack details. Additionally, latent features at different scales are quantized into discrete representations using a codebook. To learn the distribution of these discrete representations from non-crack samples, the study utilizes PixelSNAIL, an autoregressive model used for sequential modeling. In the testing stage, the model assigns low probabilities to discrete features that deviate from the non-crack distribution. These potential crack candidate features are resampled using vectors in the codebook that exhibit the highest dissimilarity. The edited representations are then fed into the decoder to generate resampled images that have the most significant differences in the crack area from the original reconstruction. Crack patterns are extracted at the pixel level by subtracting resampled images from the reconstruction. Prior knowledge that crack patterns often appear darker is leveraged to enhance the crack features. A robust classification criterion is introduced based on the probability given by the autoregressive model. Extensive experiments were conducted using images captured from stamped metal panels. The results demonstrate that the proposed technique exhibits robust performance and high accuracy.