Tool Condition Monitoring System for Competitive Aluminum Milling

Abstract

In recent years, the auto parts industry has experienced a significant transformation, transitioning from gasoline-powered vehicles to electric vehicles, influenced by the Connected, Autonomous, Shared, and Electric (CASE) technologies trends. This shift is increasing the demand for advanced components like sensors and ECUs, requiring enhanced manufacturing techniques such as die casting and machining. However, North American manufacturers face a risk in competitiveness due to must of this mechanical parts are supplied by Asian suppliers, posing risks to increase manufacturing cost related to tariffs and logistics. To stay competitive and embrace these trends, North America needs to establish a CASE manufacturing hub to localize production. Denso is a Japanese mobility supplier that has provided advanced automobile technologies, components, and systems to major manufacturers since 1949, operating in 38 countries Denso (1 10). Established in 1996, Denso México (DNMX) has grown significantly, with four plants—two in northern Mexico, one in Silao, and a recent addition in Irapuato. As of March 2023, DNMX employs over 7,000 people, making it one of the largest facilities within Denso North America and playing a key role in the North American market for CASE products (Connected, Autonomous, Shared, and Electric vehicles). To improve competitiveness in the auto-parts and support the localization of parts the strategy of DNMX is to focus on enhancing the Monozukuri spirit1. The approach involves establishing a manufacturing foundation thru integration of advance industry 4.0 strategies, including IoT, automation, and data analytics, to optimize processes and improve efficiency and quality. In the context of CASE, the emphasis is on producing essential components like aluminum-machined cases for electric parts and inverter motors. To gain a competitive advantage, there is a significant investment in advanced technologies for machining processes, aiming to ensure cost efficiency, enhance productivity, maintain quality, and extend tool life. The real-time autonomous Tool Condition Monitoring System (TCMS) is a key element of this strategy, enhanced by Artificial Intelligence (AI), which leverages machine learning to analyze real-time data, predict tool wear, and prevent potential failures. The development and deployment of the AI-driven TCMS follow the Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology, a robust framework widely adopted for data analytics projects. CRISP-DM ensures a structured approach through six phases: business understanding, where goals and objectives align with organizational strategy; data understanding, involving detailed exploration of machining and tool condition data; data preparation, including cleaning and structuring data for analysis; modeling, where machine learning algo-rithms predict tool wear and failure; evaluation, assessing model accuracy and alignment with objectives; and deployment, integrating the AI system into manufacturing processes. This methodology enhances the iterative refinement of predictive capabilities, ensuring alignment with strategic objectives and operational realities. By adopting CRISP-DM, DNMX ensures the systematic development of its AI-integrated TCMS, enhancing machining accuracy and reliability, optimizing maintenance schedules, and reducing downtime. This structured approach continuously improves the system, reinforcing DNMX’s leadership in the North American auto-parts industry and contributing to the transformation towards electric vehicles.