Advanced Computing in the Age of AI | Sunday, May 9, 2021

Noodle.ai’s FlowOps Uses Explainable AI to Streamline Production, Supply Chains 

via Shutterstock

Among the latest iterations of AI software is a category dubbed FlowOps, or Flow Operations, designed to reduce waste in manufacturing while helping to unclog vulnerable supply chains.

FlowOps allows manufacturers and other operators to manage workflows from production and inventory management to gauging product demand and maintaining quality control.

The latest FlowOps platform, unveiled March 15 (Monday) by supply chain specialist Noodle.ai, leverages capabilities such as explainable AI to address multi-trillion-dollar losses annually attributed to product defects, unscheduled manufacturing downtime and excess inventory often associated with surge demand. Explainable AI is a concept that "takes in vast amounts of data about every action in a manufacturer's operations and identifies patterns that were previously undetectable," according to the company.

The San Francisco-based startup noted growing U.S. government efforts by the Biden administration to harden supply chains, the vulnerability of which have been exposed by the COVID-19 pandemic, including efforts to add greater predictability via emerging AI tools. Those capabilities could help gauge demand, manage assets and reduce waste, much of it resulting from inefficient manufacturing workflows.

Noodle.ai’s supply chain platform includes five enterprise AI flow applications spanning asset management, quality control, production, inventory and demand workflows. The framework addresses operation “entropy” in manufacturing and supply chains that result in production delays, poor quality control and the resulting waste. Those problems can then cascade to idled supply chain capacity, supply-demand mismatches and costly downtime.

Supply chain vulnerabilities exposed by the pandemic, including widespread semiconductor shortages, prompted the Biden administration to issue an executive order in February aimed at strengthening U.S. production capacity and distribution networks. The order requires various government agencies to submit recommendations by this summer for strengthening U.S. supply chains. For example, the Commerce Department was specifically tasked with identifying risks to the chip manufacturing and IC packaging supply chains.

Despite ongoing enterprise investments in traditional tools like customer relationship management and enterprise resource planning, “operating metrics are stubbornly stuck or worsening and the pressure on planners and operators intensifies,” Noodle.ai CEO Stephen Pratt wrote in a blog post unveiling the five FlowOps applications -- Asset Flow, Quality Flow, Demand Flow, Inventory Flow and Production Flow. “One key villain should be blamed for disrupting flow: Operations entropy,” he wrote.

The result, Pratt added, is randomness, uncertainty and unpredictability on the factory floor and amid supply chains, a situation that only gets worse as complexity increases.

To cut through the fog, the company’s AI-based FlowOps platform discerns patterns in manufacturing operations, learns from those patterns, predicts scenarios and makes recommendations on each phase of production from raw materials to finished products. The workflow framework makes use of machine learning, HPC and data storage technologies, according to the company.

It also leverages explainable AI to sift through data generated by each manufacturing step to help identify “previously undetectable” workflow patterns, the company claims. The resulting recommendations are ranked based on which manufacturing and logistics decisions are deemed most critical to a workflow. The explainable AI function running on a multi-petaflop platform loads, synchronizes and analyzes time-series data, then applies ensemble models to optimize workflows. Successful results can lead to uninterrupted production lines, reduced waste and supply chains better adapted to scaling up or down to account for swings in demand.

Noodle.ai said it has invested $110 million over the last five years to developing its AI software portfolio. Among its partners are Amazon Web Services, Intel Corp. and Mitsubishi Corp. Dell Technologies Capital is an investor.

Supply chain management is seen as a prime candidate for AI adoption, according to a recent Gartner forecast which predicts that half of distribution specialists will invest in AI. “Companies will continue to invest in applications that embed, augment or apply AI and advanced analytics tools,” the January report concluded. “This may be to address foundational areas such as data quality or connecting disparate silos, or strategic objectives such as migrating to more automated, resilient and smarter applications,” the report added.

 

About the author: George Leopold

George Leopold has written about science and technology for more than 30 years, focusing on electronics and aerospace technology. He previously served as executive editor of Electronic Engineering Times. Leopold is the author of "Calculated Risk: The Supersonic Life and Times of Gus Grissom" (Purdue University Press, 2016).

Add a Comment

EnterpriseAI