Feasibility study for an automated Machine Learning analytics platform for better process control of artificial retina in-space manufacturing

LambdaVision is developing an artificial retina to restore vision to the millions of people blinded by retinal degenerative diseases. The artificial retina is manufactured using a layer-by-layer (LBL) assembly technique, which results in an artificial retina thin film that contains hundreds of oriented layers of the light-activated protein, bacteriorhodopsin.nbsp;nbsp;The LBL process is subject to the effects of gravity, which leads to inefficient and homogenous layering. As such, the artificial retina is example of a drug product that may benefit from production in low earth orbit (LEO).nbsp;Through Space Tangorsquo;s CubeLabTMnbsp;system and seven research flights to the ISS, LambdaVision has established proof-of-concept for artificial retina manufacturing in LEO. Thisnbsp;Phase I SBIRnbsp; proposal will build on results obtained from our microgravity work and parallel terrestrial efforts to establish an informed data driven process via machine learning in order to efficiently and effectively leverage each experiment conducted in microgravity.nbsp;nbsp;nbsp;LambdaVision will partner with Dr. Ioana Cozmuta and her team at G-SPACE to use artificial intelligence (AI) and machine learning (ML) tools to systematically de-risk the key parameters required to achieve a viable microgravity product.nbsp;nbsp;Through G-SPACErsquo;s proprietary platform that uses data science, ML, and software automation tools, we will evaluate artificial retina thin films generated under various conditions to help establish real time quality and process control during the manufacturing process. These data will help to reduce the current trial and error approaches in LBL manufacturing and remove uncertainty from the experiments by using proprietary algorithms established via decades of microgravity research. By standardizing the experiments and achieving the process controls for experimentation we will maximize the data collected, reduce costs, and accelerate the timeline to commercialization.nbsp;nbsp;