Scholarlink

The emergence of Unmanned Vehicles (UxVs) has revolutionized various industries, offering unprecedented capabilities in areas such as surveillance, logistics, and environmental monitoring. As UxVs become increasingly integral to critical operations, the reliability of their components, particularly semiconductor chips, becomes paramount. Ensuring chip reliability is crucial for maintaining the overall performance and safety of UxVs, particularly in critical systems such as navigation, obstacle detection, environmental sensing, and data transmission. Reliability failures can be exploited to breach security, leading to unauthorized access or data corruption. In recent years, advancements in semiconductor technology have increased the vulnerability of semiconductors to reliability issues caused by deterioration over time, such as transistor aging. Aging related failures are accelerated by rising temperature, particularly emphasized in advanced technologies due to the increase thermal density. Resolving such reliability issues becomes challenging due to the complex interplay of physical mechanisms, environmental conditions, and compute workloads, each requiring adaptive mitigation strategies. This paper introduces a novel predictive maintenance system for detecting and mitigating reliability failures in UxVs. To the best of our knowledge, this is the first predictive maintenance system introduced specifically for IC reliability. The proposed system can analyze multi-dimensional sensor data to provide comprehensive and robust predictive maintenance, dynamically addressing evolving aging faults. The proposed system consists of three subsystems: A set of sensors that can identify evolving failures before they manifest in the crucial functional systems of the UxV; a set of maintenance agents that can mitigate various types of progressing failures; and a control subsystem that orchestrates the sensors and the maintenance agents. Our preliminary simulations suggests that the proposed system has the potential to mitigate such reliability failures.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predictive Maintenance System for Enhancing Chip Reliability and Resiliency in UxVs

  • 会议论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年

查看全文
  • Freddy Gabbay

摘要

The emergence of Unmanned Vehicles (UxVs) has revolutionized various industries, offering unprecedented capabilities in areas such as surveillance, logistics, and environmental monitoring. As UxVs become increasingly integral to critical operations, the reliability of their components, particularly semiconductor chips, becomes paramount. Ensuring chip reliability is crucial for maintaining the overall performance and safety of UxVs, particularly in critical systems such as navigation, obstacle detection, environmental sensing, and data transmission. Reliability failures can be exploited to breach security, leading to unauthorized access or data corruption. In recent years, advancements in semiconductor technology have increased the vulnerability of semiconductors to reliability issues caused by deterioration over time, such as transistor aging. Aging related failures are accelerated by rising temperature, particularly emphasized in advanced technologies due to the increase thermal density. Resolving such reliability issues becomes challenging due to the complex interplay of physical mechanisms, environmental conditions, and compute workloads, each requiring adaptive mitigation strategies. This paper introduces a novel predictive maintenance system for detecting and mitigating reliability failures in UxVs. To the best of our knowledge, this is the first predictive maintenance system introduced specifically for IC reliability. The proposed system can analyze multi-dimensional sensor data to provide comprehensive and robust predictive maintenance, dynamically addressing evolving aging faults. The proposed system consists of three subsystems: A set of sensors that can identify evolving failures before they manifest in the crucial functional systems of the UxV; a set of maintenance agents that can mitigate various types of progressing failures; and a control subsystem that orchestrates the sensors and the maintenance agents. Our preliminary simulations suggests that the proposed system has the potential to mitigate such reliability failures.