Robust Metal Foreign Object Detection for Wireless Charging in Mobile Systems

Wireless charging, which transmits energy via invisible electromagnetic fields without cumbersome cables, is transforming how we power mobile systems. From traditional mobile devices like phones to emerging autonomous mobile machines like drones, it enables convenient ubiquitous charging experiences (e.g., in-car charging) and even irreplaceable charging operations (e.g., drone's auto-docking at base stations, where precise alignment for wired charging is impractical). With a rapid growing of compatible devices, wireless charging is evolving into vital urban infrastructure, seamlessly integrated into various places.However, this innovation brings a hidden issue: metal foreign objects, e.g., SIM-card ejectors, keys, coins, or industrial metal debris, can inadvertently enter the charging zone. These intruders absorb electromagnetic energy, inducing eddy currents that trigger rapid overheating, posing severe potential risks of fires, burns, damages to devices or batteries. The risk will intensify in interconnected setups, like in-car phone charging or drone workstations, where a single risk could cascade into catastrophe.Consequently, wireless charging protocols mandate metal foreign object detection. For example, objects must be detected before reaching a 70°C thermal threshold in the dominant Qi protocols (beyond which risks may occur) and charging should halt instantly. Alarmingly, our preliminary evaluations of commercial chargers for both mobile phones and drones expose critical flaws: most fail to detect metal foreign objects reliably. This stems from the inherent limitation and unreliability of existing detection methods, hampered by device interference and object properties. Therefore, existing methods prioritize false-alarm avoidance, resulting in frequent missed detections.In this project, we unleash the untapped potential of exploiting specialized fingerprints of metal objects in the charging zone: their coupling with the charging electromagnetic field selectively suppresses high-frequency signal harmonics, offering a distinctive, reliable signature for detection. However, realizing this design encounters two key challenges. First, capturing these subtle fingerprints demands innovative and robust sensing. In this project, we propose a specialized sensing circuit and software pipeline optimized to precisely harvest them. Second, variations across objects (e.g., materials and sizes) and protocols (e.g., for traditional or emerging mobile devices) introduce inevitable distinctions. We further propose effective countermeasures for seamless generalization.Project's significance is profound. It delivers a remarkable leap in wireless charging safety, safeguarding everyday use and unattended scenarios against risks. By delving into signal-level insights, we enhance the interpretability and reliability of our solution.Overall, the idea to exploit metal-inspired signatures for robust foreign object detection is of essential novelty. The proposed techniques are also innovative.