Machine learning (ML) and artificial intelligence (AI) in industrial settings have certainly been getting a lot of buzz, but their adoption in process manufacturing has been limited, according to Oden Technologies (New York, NY). The company, which describes itself as a developer of intelligent industrial automation, hopes to change that dynamic with the introduction of the industry’s first end-to-end ML and AI framework for manufacturing.
A one-size-fits-all approach to ML and AI will never deliver on the full potential of those technologies, said Deepak Turaga, Vice President of Data Science at Oden Technologies, in a press release distributed today. The complexity and specificity of manufacturing processes have traditionally demanded heavily customized solutions developed by internal data science teams. Oden’s approach is to “provide customers with the best-in-breed foundational ML and AI applications tailored for manufacturing, and the framework tools to easily extend and adapt them to their specific requirements and processes,” said Turaga. For example, the open framework features algorithms and data science tools for extrusion and injection molding processes.
Oden’s production-ready framework integrates ML algorithms and data science tools with both structured and unstructured data from machines, operator inputs, quality assurance, work orders, environmental monitors and product specs. The system allows manufacturers to take preventive actions to avoid machine failure, eliminate waste and optimize production in real time.
The hybrid cloud and edge design reduces latency and dependence on connectivity—it delivers the power of the cloud without compromising the requirements of mission-critical applications, said Oden.
Customers using the existing Oden platform, which the new framework will augment, typically report a 20% percent increase in monthly output and 50% decrease in total scrap, resulting in millions of dollars in savings and additional revenue each year, according to Oden. The open and extensible nature of the new ML and AI framework is expected to significantly accelerate these gains.
Beyond improving current production processes and operational efficiency, ML can be used to capture both implicit and explicit process know-how, allowing manufacturers to leverage the system in bridging their talent skills gap, said Oden.
“Systems like this help to institutionalize and operationalize domain knowledge, so it can be preserved by the manufacturer and shared more easily with others for training,” Willem Sundblad, co-founder and CEO, explained to PlasticsToday. “The existing Oden platform already helps to close the skills gap because it allows users to solve problems faster, which enables them to move faster with smaller teams. This is crucial since plastics manufacturers are struggling to fill highly skilled roles. They need to make their current workforce as productive as possible and focus on value-adding tasks,” Sundblad said.
The new ML/AI framework will further close this gap, he claims, because it will allow manufacturers to perform analysis and forecasting at an unprecedented scope. “This is extremely valuable technology for manufacturers, but building it in-house is often out of reach for them because it is expensive and attracting the talent to do it is difficult. That’s where solutions like Oden come in,” said Sundblad.
Image courtesy Olivier Le Moal/Adobe Stock.