For factories that are packed with workers, identifying missteps and bottlenecks can be quite challenging. In this regard, AI-based video analytics can come in handy
Needless to say, intelligence and automation are popular concepts among manufacturers these days. Yet for manufacturers in certain industries such as EMS (electronics manufacturing service), garment and shoe manufacturing, heavy manual labor is still required to get things done. For factories
that are packed with workers, identifying missteps and bottlenecks can be quite challenging. In this regard, AI-based
video analytics can come in handy, helping manufacturers improve processes and productivity.
is one company that offers AI-based analytics for applications in manufacturing. The start-up, based in Hong Kong and now with offices in Taiwan and the U.S., was founded by Ken Law and Hauman Chow, both former Google employees with relevant factory experiences, knowing first-hand about the challenges and difficulties that factory operators are faced with. The two then embarked on a journey to develop solutions that address manufacturers’ pain points.
According to Law, who is also the company’s CEO, a slowdown in production is usually due to a particular bottleneck on the production line, which is characterized by an upstream-downstream relationship where one worker passes a particular product or task to another. If one person is slow in handling a particular task, the entire line can be affected.
Typically, what the manufacturer does is hire an industrial engineer (IE) to oversee production and make sure operations are running smoothly. Yet for labor-intensive industries like electronics, garments and shoe manufacturing where each production line can easily have 50, 80 or even 100 workers, pinpointing bottlenecks or problems can be quite difficult.
“A lot of IEs today use stop watches, pencil and paper to get their jobs done, which is quite insufficient,” Law said.
Analyzing root cause in real-time
This is where PowerArena’s AI solution comes into play. Its edge server-based AI takes video streams from IP cameras and helps find out what the root cause of a problem might be. For example, the AI can identify a worker who spends longer time working on a particular task than others. A report is then generated to help the IE rectify the issue accordingly, for example providing the necessary support or assistance to that worker.
In a similar vein, PowerArena’s AI can detect impending machine failure and other machine issues. For example, machines that are idling can lower the factory’s overall equipment effectiveness or OEE. In this regard. PowerArena’s AI can detect an idling machine and help operators determine why the machine is not running – is it because the machine has failed or is it because the upstream workers haven’t completed their tasks yet. If the latter is the case, the IE can make relevant adjustments to improve the process.
“Using deep learning AI technology, PowerArena turns typical webcams into powerful sensors which find out the bottleneck of manufacturing operations and analyze the root cause in real-time. It helps you to gain factory floor visibility, increase OEE and establish the data foundation for Industry 4.0
,” Law said.
According to Law, PowerArena’s AI has proven quite effective in helping clients improve productivity. “One of our clients used our solution for one month to make improvements on processes, and they found out that by using our solution, productivity increased by more than 5 percent,” he said.
In terms of distribution, PowerArena either goes directly to the end user or works with partners such as ADLINK and Advantech in pushing its technology. As for geographic regions, PowerArenan focuses on markets where manufacturing is more labor-intensive. “These include China, Taiwan and Southeast Asia markets, but PowerArena is also going to invest more resources in the U.S., because their labor cost is really expensive compared to Asia,” Law said.