Software automation It is a robotics company that wants to make T-shirts. “We hope to produce 1 billion T-shirts every year in the United States, all on-demand,” said Palaniswamy Rajan, CEO of SoftWear.
In 2012, the company The contract between the Georgia Institute of Technology Advanced Technology Development Center and Darpa. Two years later, a prototype started running. By 2017, start to develop a production line that can mass produce shirts.In the same year, the company reached an agreement with a Chinese clothing manufacturer to set up a large Arkansas production facilityHowever, the deal failed and SoftWear is now focusing on opening its own garment factory.
The time taken to reach this point is not surprising.It turns out that the machine is proficient in many steps of making clothes, from Printed textiles arrive Cut fabric with Folding and packaging finished garments.
But as we all know, sewing is difficult to automate because textiles gather and stretch during processing. When the fabric passes through the sewing machine, human hands are very good at keeping the fabric in order. Robots are usually not dexterous enough to handle tasks.
SoftWear’s robot overcomes these obstacles. They can make a T-shirt. But making them as cheap as human workers in places like China or Guatemala, where workers earn only a small part of what they earn in the United States, will be a challenge, said Sheng Lu, a professor of fashion and apparel research at South China University of Technology. . Delaware.
SoftWear calls its robotic system Sewbots. They are basically elaborate workbenches that pair sewing machines with complex sensors. The company pays great attention to details, but the basic principles are as follows: The fabric is cut into small pieces and becomes part of the shirt: front, back and sleeves. These parts are loaded onto a work line, where a complex vacuum system stretches and moves the material instead of a person pushing the fabric through the sewing machine. The camera tracks the lines in each panel, allowing the system to make adjustments as the garment is made.
But no two batches of cotton are exactly the same, and they often vary from harvest to harvest; changes in fabrics and dyes further complicate the problem. Each change may require recalibration of the system and interruption of operation, and SoftWear must train its machine to respond accordingly. “The biggest challenge we faced when setting up a production system was the requirement to be able to run 24/7 at high speed and with a quality of more than 98%,” Rajan said.
Apparel factories produce more than 20 billion T-shirts each year, most of which are outside the United States. In order for a T-shirt to be made in the United States, it must be cheaper than imported. But eliminating transportation costs and import tariffs is not enough to cover the cost of sewing clothes for American workers. According to the US Bureau of Labor Statistics, the average annual income of sewing machine operators in the United States is less than $28,000. This is about $13.50 per hour-much higher than countries that currently produce many T-shirts. Delaware professor Lu said that the wages for such jobs in China are about one-third of the wages in the United States, while the wages in Guatemala are less than one-fifth of the wages in the United States.
Focusing on T-shirts allows SoftWear to avoid another problem with automated sewing systems: switching from one garment to another. A skilled human team might one day sew short-sleeved men’s shirts and then women’s jeans. This transition is more challenging for robots. The way that cotton polo shirts are sewn is quite different from the way that polyester pants are constructed. It is complicated and costly to develop new working threads and sew different stitches for different fabric cuts. Once a T-shirt production line is established, it is difficult to quickly reconfigure sewing robots to produce other things.
Since the initial financing, SoftWear has raised US$30 million in venture capital and grants, including a US$2 million grant from the Walmart Foundation. Rajan said that it would take tens of millions of dollars to produce 1 billion T-shirts each year. To achieve this goal, the company needs multiple facilities, each with its own sewing robots and skilled workers to maintain them. Rajan said that the Sewbot production line can make a T-shirt every 50 seconds. At this rate, if it runs continuously, a production line can produce more than 620,000 T-shirts per year-which means that 1,607 Sewbots are required to work continuously to reach 1 billion in a year. Rajan said a more realistic number is close to 2,000; so far, the company has only produced less than 50 units.
Robots inevitably arouse people’s suspicion about replacing people and disrupting work. Rajan admits that SoftWear will hire fewer people than traditional T-shirt manufacturers, but he believes that his company will create higher-paying jobs for people maintaining machines. “You want to develop the workforce, you want to train the workforce,” he said. “Our goal is to have a skilled labor force and fast and flexible production.”