If knowledge collected from manufacturing course of will be properly concatenated, then issues affecting manufacturing effectivity will be rapidly noticed via massive knowledge evaluation, in keeping with Alex Hsu, CEO of TAO Info, a Taiwan startup devoted to providing knowledge analytics to assist producers diagnose their manufacturing strains and optimize manufacturing course of.
In reaching sensible manufacturing, Hsu stated, knowledge assortment and evaluation is essential preparatory work, however Taiwan producers nonetheless have a lot room for enchancment on this facet as lots of them are simply beginning to use knowledge evaluation for single goal resembling defect picture evaluation or predictive gear upkeep.
Hsu pressured that the ultimate goal of sensible manufacturing is to not improve the efficiency of any single system however to optimize all the manufacturing course of and increase yield charges via AI-based evaluation of concatenated knowledge.
Hsu, who used to have interaction within the semiconductor sector, stated it’s a thorny downside for IC engineers to find production-affecting elements from amongst 1000’s of frontend processes, identical to searching for a needle in a haystack. But in analyzing massive knowledge, he continued, algorithms can be utilized to outline issues via an enormous quantity of parameters regarding equipment, materials, system, operator and setting and rapidly kind out key elements affecting manufacturing in accordance with correlation among the many parameters. Such a apply is aimed toward narrowing the vary of doable elements.
Hsu continued that the majority AI or neural community evaluation applied sciences can now be utilized to cope with correlation relatively than the cause-and-effect relation between course of parameters and manufacturing issues, and due to this fact his firm can use a search engine to kind out the almost certainly related parameters by conducting a correlation-based sequencing of all of the parameters, permitting engineers to rapidly deal with issues.
Manufacturers may incorporate AI to attain sensible product traceability along with sensible manufacturing, Hsu famous, furthering that many shopper electronics distributors together with Lenovo, Huawei and Dell have requested their supporting companions to actualize “digital twin” purposes, in order to concatenate all of the manufacturing processes of all the provide chains.
TAO Info CEO Alex Hsu
Photo: Chloe Liao, Digitimes, May 2019
