CreateMe’s pitch marks a meaningful shift in apparel automation: the company is trying to move garment production from sewing to robotic adhesive bonding, with Physical AI providing the perception and control needed to handle fabric as a deformable, stateful material rather than a rigid part. That matters now because textiles have remained one of the last large manufacturing categories that industrial robots have not been able to fully absorb. If Myers is right, the bottleneck is no longer whether robots can move fast enough, but whether they can understand what the fabric is doing moment to moment and respond before the material drifts out of spec.
The technical change is not just swapping one joining method for another. Sewing assumes a human or machine can mechanically manage layers of fabric through a needle-based process, but CreateMe is describing a line built around single-sided access and adhesive bonding. That architecture reduces the need for two-sided manipulation and can simplify the robot’s physical interaction with the garment. The harder part is the control problem: fabric stretches, wrinkles, and shifts unpredictably, so the machine has to estimate real-time material state well enough to place bonds accurately and consistently. In that sense, Physical AI is doing the work conventional automation could not—closing the loop between perception, material behavior, and actuation on a deformable substrate.
That is also why the interview points to a broader industrial signal. If a bonding-based process can be made reliable, it could reorder line design in sectors where labor has remained deeply embedded in repetitive joining tasks. Apparel is the obvious first market, but the same deformable-material challenge exists in automotive interiors, medical textiles, and some composite workflows. CreateMe’s positioning suggests it sees the near-term opportunity in categories where high mix, frequent style changes, and labor intensity make automation economically painful today. Bonding, in theory, becomes attractive when it can cut manual handling, reduce rework, and compress cycle times enough to justify the capital expense.
The supply-chain implications are as important as the robotics stack. A shift from stitching to bonding changes more than the factory floor; it changes consumables, quality assurance, line logistics, and purchasing. Adhesives introduce their own procurement and storage requirements, and any production line built around them inherits new dependencies on chemistry, cure behavior, shelf life, and environmental stability. That raises the stakes for cost modeling. Labor savings only matter if adhesive costs, scrap rates, maintenance, and downtime do not erase the benefit. For buyers, the comparison will be less about whether a robot can bond fabric and more about whether the full system delivers lower unit cost at acceptable quality across production runs.
This is where the feasibility question gets concrete. CreateMe’s model needs to prove that bond durability holds under real-world wear, washing, and handling, and that performance does not degrade as fabrics, finishes, and product types vary. It also has to show that material-state estimation is robust enough for production, not just a controlled pilot. Deformable materials are notoriously sensitive to small variations, which means even a strong demo can fail at scale if the system cannot adapt to feed variability, humidity, or upstream differences in textile lots. Integration matters too: manufacturers are not starting from a blank slate, so any deployment has to fit existing supply-chain flows and line constraints rather than demanding a full plant redesign.
For that reason, the most important milestones are operational, not rhetorical. The first is repeatable bond quality at target cycle times across multiple fabric types. The second is a credible total-cost-of-ownership comparison against sewing-based lines, including labor, adhesives, maintenance, and scrap. The third is evidence that the system can be inserted into real production environments without creating new bottlenecks elsewhere in the factory or supplier network. If CreateMe can clear those hurdles, bonding may move from an interesting workaround to a serious manufacturing paradigm for deformable goods. If not, the gap between a promising Physical AI stack and production-scale economics will remain wide.
