In 1975, a young Engineer at Kodak assembled a prototype digital camera. In comparison to a film camera, it was primitive, the bulky camera produced very poor quality images. The Kodak management failed to see latent disruptive potential in it and remained busy with their on-going profitable film camera business. But Sony leadership saw the hidden potential and kept financing research in improving digital imaging technology. Within less than 20 years, the once primitive digital camera grew as a disruptive force and eventually succeeded to disrupt film camera industry-making American icon Kodak bankrupt. At the early stage, often-high potential technology emerges in primitive form with a misleading impression about future prospects. Often incumbent firms having business around matured technology core overlook the potential, suffering major setback at a later stage. It appears that today's sewing robots are primitive in comparison to the fabric handling capacity of human fingers-powered by high performing hand-eye coordination. But could it grow as a strong substitute to human hands, causing disruption to global sourcing of apparels?
Due to the labour cost differential, the apparel industry has been mobile. It has been moving over the last couple of centuries to take advantage of low-cost labour from across the world. This cost-sensitive industry has migrated from western countries to far away destinations like Bangladesh, Vietnam, and Cambodia. But will it stay in these countries? Or, is it the last leg of the journey before getting back to pavilion? There has always been a prediction, "Soon, retailers may not have to wait for long periods for a clothing shipment to reach their country, thanks to the SAM-1000, an automatic sewing machine that will operate completely on its own, without a human operator." The question could be: how soon? It's time to predict the growth pattern of the underlying technologies and interpret likely implications as a substitute to human labour.
In the value chain of apparel making, automating fabric handling during sewing has been found to be a significant hurdle. The nature of the fabric itself has been the cause of a longstanding problem with automating sewing. It's flexible, it stretches, it warps, and it folds. Stitching two pieces of such fabric together involves aligning the material correctly to the sewing head, feeding it through and constantly adjusting the fabric to prevent it slipping and buckling, while all the time keeping the stitches neat and the thread at the right tension. This has been a very tricky barrier to realising the objective to make a machine in which the fabric goes in one end and finished garments, such as jeans and T-shirts, come out the other. It has been found that nimble fingers, with close visual feedbacks, invariably prove better at this than cogs, wheels and servo motors.
For dynamically aligning pieces of fabrics in feeding to swing head, very precise and rapidly varying hand-eye coordination is essential. Poor camera resolution and taking seconds to process images in providing feedback were failing to enable the machine to mimic the simultaneous use of hands and eyes of workers, requiring the information our eyes perceive (visual perception) to guide our hands to carry out the movement of fabric. But it appears that the recent developments of very high-resolution camera sensors, powerful, but affordable high performing computing chips, and high performing computational algorithms are about to overcome this barrier. It's being reported that success has been attained in developing a specialised camera capable of capturing more than 1,000 frames per second, and a set of image-processing algorithms to detect, on each frame, where the threads are in real-time, without slowing down the speed. Such high speed, also high-resolution vivid image frames allow the computer to pick out individual threads in the fabric. As a consequence, any distortion to the fabric made by each punch of the needle can be measured extremely accurately. As reported by the Economist, these measurements are vital achievement in allowing the "feed dog", which gently pulls the fabric through the machine, to make constant tiny adjustments to keep things smooth and even. This success appears to be sufficient enough to mimic required human hand-eye coordination in handling fabrics-making sewing steps human-free.
Upon patenting this machine vision guided fabric feeding technology, a start-up in Georgia of the USA has developed a highly calibrated machine vision innovations to watch and analyse fabric-succeeding in detecting distortions and robotically adjusting the fabric, while feeding to sewing head. In the demonstration plant, working across a 70-foot long T-shirt production line, the robot performs each task, including cutting, sewing a seam, adding a sleeve, and quality inspection. And at each step of the way, the computer vision guides the fabric. It's being reported that "using this high caliber machine vision and real-time analysis, the robotics then continually manipulate and adjust the fabric to be properly arranged."
By leveraging this automated sewing technology, a Chinese clothing manufacturer, who produces clothing for Adidas and Armani, has opened its newest factory in Arkansas, not China. In this new factory, three to five people, as opposed to 10, will work each of the 21 robotic production lines, resulting in 50-70 per cent less labour requirement. Moreover, the robots will also increase production. A human sewing line produces 669 T-shirts in eight hours, compared to the robots at 1,142 t-shirts-- a 71 per cent increase in production. As a result, the labour cost of production of apparel in US plants will rapidly fall. As reported, in Bangladesh the labour cost to produce a denim shirt is about $0.22. If made by U.S. workers, that cost jumps to $7.47, but with a robotic production line, it's just $0.33 per T-shirt." It's claimed that this robotic technology working under the guidance of a single human handler will be producing as many shirts per hour as about 17 humans-making machine cheaper than the least costly apparel making labour of Bangladesh or Cambodia.
This type of technological advancement would have a major impact on the Asian countries that employ low wage workers for meeting the demand of western apparel and footwear buyers. The estimate made by the International Labour Organisation that robots will replace 64 per cent of textile, clothing, and footwear workers in Indonesia, 86 per cent in Vietnam, and 88 per cent in Cambodia is worth taking into consideration in guiding our decision. It's highly important for these countries to keep monitoring and predicting such technological advancements and assessing likely implications on their major industrial jobs as well as exports.
M Rokonuzzaman Ph.D is academic and researcher on technology, innovation and policy. [email protected]