Manufacturing AUTOMATION

News Artificial Intelligence Technology
Element AI partners with Japanese automotive parts manufacturer


November 5, 2019
By Manufacturing AUTOMATION

Topics

Element AI, a global developer of artificial intelligence-powered (AI) software services, has entered into a strategic relationship with Aisin Seiki Co. of Japan with the goal of encouraging the use of computer vision in automotive manufacturing operations.

Manufacturers are rapidly adopting AI to facilitate how they identify and verify anomalies that can compromise quality during production processes. Recommendations made by a machine learning (ML) algorithm without an easy-to-follow explainability path often causes distrust with users and non-compliance with customers.

Element AI includes explainability features – what it dubs Explainable AI – in many of its AI software products and is developing AI solutions to help address the three biggest factors facing manufacturers today: the projected global skills gap from lost skills and knowledge transfer, high velocity and volume of data generated due to Industry 4.0 and IoT connected devices, and rapidly evolving and changeable consumer demand.

“Visually tracking and watching for any possible anomaly requires a high level of computer vision accuracy,” says Katsuaki Takahashi, group manager of Aisin Seiki. “Incorporating Explainable AI in our factories is an example of our commitment to assuring confidence in the quality and reliability of our processes and products.”

All of these data and learning challenges require sophisticated machine learning applications to analyze, learn, recognize and predict – with explained steps – how the algorithm arrived at its recommendation or analysis.

Karthik Ramakrishnan, head of industry solutions at Element AI, says, “We are excited to begin this collaboration between Aisin Seiki Group and our Element AI development team to design and implement explainability protocols for more accurate and reliable anomaly detection, which will ensure confidence in product quality.”