Yehoram Yosubash, Hsiao-Lung Chang, Louis Dagenais, Nate Spiker, Chris Aitken, Stu Beale, Brooks Automation, Chelmsford, MA
Since its founding in 1978, Brooks Automation has been a leading automation provider and trusted partner to the global manufacturing industry. The first wafer-handling MagnaTran^® vacuum robot was introduced in 1994. More than 35,000 have since been shipped and are found in manufacturing facilities worldwide transferring substrates to produce highly advanced logic and memory chips, LEDs and optoelectronics, hard-disk drives, and optical lenses and components.
Traditional wafer-handling robots use belts and gears to transfer rotary servomotor motion of robot joints to an S-curve motion profile at the end effector. The Brooks Automation MagnaTran^® robot is the world’s first direct-drive vacuum robot that uses an innovative in-vacuum rotor and shaft-coupled encoders to deliver infinite rotation, precision substrate placement, and fast substrate-swapping. Complex motion controls have also been developed to harness the direct-drive torque.
The semiconductor industry aggressively demands improved throughput, high yield, and energy efficiency, as well as a lower cost of ownership. The MagnaTran^® LEAP^™ robot provides these solutions with best-in-class performance and longevity for leading edge Memory and Logic chip production.
A ground-up redesign of the robot’s hardware and its manufacturing process further strengthens and differentiates performance and capabilities over alternative solutions for leading-edge applications. Statistical Process Control (SPC) and automated monitoring are utilized during the build and test environment. Re-written software paired with on-board diagnostic and motion analysis bolsters the MagnaTran^®’s hardware capabilities.
Template move motion analysis is an SPC-like monitoring and alerting capability that can detect small variations in movement trajectories for specific station-to-station paths over the lifetime of individual robots. Like traditional SPC, template move motion analysis collects initial benchmark motion performance data of the robot for specific station-to-station motion paths to capture the normal range of variability in movement trajectories. This, in turn, allows the robot to monitor its performance and alert host software whenever the robot starts to deviate outside of its established level of variability. Monitoring like this can detect more subtle changes in the performance of individual robots, such as peak torque and acceleration.