AI Innovation Program

The current fixed schedule and reactive maintenance approach for military vehicles results in unexpected failures and prolonged downtimes, negatively impacting maintenance teams, operational units, and supply chain management. These issues lead to increased costs, reduced vehicle availability, and heightened safety risks. Additionally, the Army's structured four-level maintenance system falls short in predicting and preventing such problems. There is a pressing need for a digital twin-based predictive maintenance system that leverages real-time data and AI to anticipate maintenance needs and potential failures, thereby enhancing operational readiness, reducing costs, improving safety, and optimizing resource allocation.

The factory is equipped with sensors to monitor crucial equipment data like temperatures, speeds, vibrations, and energy consumption. Currently, Ipack is developing a Central Data Acquisition System (CDAS) and Manufacturing Execution System (MES) to aggregate this sensor data using historian applications, driven internally. Now, they seek external support to integrate AI solutions for predictive maintenance. This aims to leverage CDAS and MES data to minimize breakdowns and unplanned maintenance events, addressing challenges posed by outdated maintenance practices in manufacturing. The goal is to have an AI-driven predictive maintenance system owned by, trained by and sitting at Ipack to enhance equipment performance, reduce downtime, and optimize operational costs, thereby improving overall plant productivity and efficiency.

Managing breakbulk cargo presents significant challenges due to its requirement for individual handling, which leads to issues such as unloading fewer goods than expected, sending cargo to the incorrect port destination, delivering cargo to the wrong recipient , and receiving more cargo than planned. These problems are further complicated by human errors during the unloading process. To address these challenges, we seek an automated, AI-based, and secure solution that ensures real-time counting, seamless integration with existing systems to enhance cargo visibility, reduce errors, and improve overall logistics efficiency, leading to faster, more reliable deliveries.

Enhancing the accuracy and efficiency of tile shade matching remains a challenge in manufacturing. Currently, this process relies on visual inspection and manual adjustments to glaze or digital printing colors, which can be time-consuming, inconsistent, and lead to production waste. We seek an AI-powered solution that integrates AI and robotics to optimize real-time color adjustments during tile production. This solution should analyze tile shades under specific lighting conditions, accurately identify discrepancies, and automatically adjust color percentages and other machine parameters to achieve a precise match. This will improve shade matching accuracy, reduce manual intervention, minimize waste, and increase production speed and output.