The pharmaceutical manufacturing industry has long struggled with monitoring the properties of a dry mix, a critical step within the production of medication and chemical compounds. Currently, two noninvasive characterization methods are typically used: a sample is either imaged and individual particles are counted, or researchers use scattered light to estimate the particle size distribution (PSD). The former is time-intensive and ends in more waste, making the latter a more attractive option.
In recent years, MIT engineers and researchers have developed a physics- and machine learning-based scattered light approach that has been shown to enhance pharmaceutical pill and powder manufacturing processes, increasing efficiency and accuracy and leading to fewer defective product batches. A brand new open-access paper, “Noninvasive estimation of powder size distribution from a single speckle image”, available within the journal , extends this work and introduces an excellent faster approach.
“Understanding the behavior of scattered light is one of the vital topics in optics,” says Qihang Zhang PhD '23, associate researcher at Tsinghua University. “By making advances within the evaluation of scattered light, we’ve also invented a useful gizmo for the pharmaceutical industry. Locating the weak point and solving it by studying the elemental rule is essentially the most exciting thing for the research team.”
The paper proposes a brand new PSD estimation method based on pupil technique that reduces the variety of frames needed for evaluation. “Our learning-based model can estimate the powder size distribution from a single speckle snapshot, reducing the reconstruction time from 15 seconds to simply 0.25 seconds,” the researchers explain.
“Our essential contribution on this work is the 60-fold acceleration of a particle size detection method through a joint optimization of algorithm and hardware,” says Zhang. “This high-speed probe can detect size evolution in fast dynamic systems and provides a platform for studying process models within the pharmaceutical industry, including drying, mixing and mixing.”
The technique provides a low-cost, non-invasive particle size probe by collecting light backscattered from powder surfaces. The compact and portable prototype is compatible with most drying systems available on the market so long as an commentary window is out there. This online measurement approach may help control manufacturing processes, thereby improving efficiency and product quality. In addition, the previous lack of online monitoring prevented the systematic investigation of dynamic models in manufacturing processes. This probe could provide a brand new platform for conducting series research and modeling of particle size evolution.
This work, a successful collaboration between physicists and engineers, was carried out as a part of the MIT Takeda Program. The collaborators come from three MIT departments: mechanical engineering, chemical engineering, and electrical engineering and computer science. George Barbastathis, professor of mechanical engineering at MIT, is the lead creator of the paper.
