In recent years, energy-storage battery technologies and related applications have advanced rapidly, with associated safety concerns becoming increasingly prominent. According to Beijing Institute of Technology (BIT), a team led by Prof. Hao-Sen Chen at the Institute of Advanced Structure Technology has independently developed a non-intrusive implantable intelligent sensing systemfor individual cells. The system enables long-term, stable and accurate measurement and wireless transmission of internal battery signals, allowing early failure diagnosis and warning, thereby markedly improving battery safety and long-term operational stability. The results were published in Nature.
Conventional external sensing struggles to capture early, high-fidelity risk signals from inside cells. Implantable internal sensing has thus emerged as a hot topic; however, several schemes proposed in Europe and the United States still face technical bottlenecks, including compromising the sealed cell structure, electromagnetic shielding that restricts signal transmission, insufficient long-term stability, and poor compatibility with industrial manufacturing.
After more than a decade of cross-disciplinary work, Prof. Chen’s team has put forward an innovative “China solution” featuring four key technical attributes:
Corrosion-resistant, accurate sensing (“measure well”): by developing 50-μm-thick thin-film sensors resistant to chemical/electrochemical corrosion, the scheme reconciles the long-life requirement for implanted sensors with the corrosive electrochemical environment.
Non-intrusive implantation compatible with manufacturing (“embed well”): an implantation process compatible with industrial cell assembly resolves the tension between sensor implantation and stable service over the full cell life.
Transmission through shielding (“transmit well”): a miniaturized communication chip using carrier-based/power-line communication sends internal signals out through the tabs despite the cell can’s electromagnetic shielding.
Intelligent early warning (“use well”): data-driven analysis models built on long-term internal sensing enable preliminary early warning of incipient cell failures.
Based on this architecture, the team designed a miniaturized, low-power implantable sensing system that precisely senses and wirelessly transmits internal temperature and strain signals in lithium-ion cells. In a commercial 100-Ah prismatic LiFePO₄(LFP) cell, the system exhibits ultralow power consumption; prismatic cells integrated with the sensing system show high cycling stability over 1,000 cycles with 93.74% capacity retention. Leveraging predictive models together with internal strain signals, the system localizes electrode fracture positions. With the implanted temperature sensor and an internal short-circuit (ISC) triggering technique, abnormal internal temperature and strain signatures can be identified at an early stage, thereby enhancing Li-ion battery safety.
Looking ahead, the team will focus on: (i) refining dedicated scientific instrumentation to provide non-destructive research tools for opening the “black box” of cell internals; (ii) constructing digital-twin batteries and, together with in-house multi-scale simulation software, achieving accurate state prediction; and (iii) engineering inherently safe batteries by integrating intelligent sensing, computational software and AI algorithms to realize “early warning and early intervention.” These advances will promote real-world deployment of intelligent sensing for energy-storage batteries in grid-scale storage plants and electric vehicles, accelerating progress toward inherent battery safety.