This paper proposes an innovative automated multiphysics Microelectromechanical systems (MEMS) co-optimization platform integrating the embedded reduced-order modeling offered by Coventor MEMS+ with MATLAB’s numerical computing environment, addressing critical limitations in conventional design workflows. Our framework demonstrates three key advances. First, the platform enables systematic co-optimization of structural parameters (comb geometries and suspension beams) through constrained design space exploration, achieving an order-of-magnitude improvement in sensor linearity while maintaining baseline sensitivity as validated by accelerometer case studies. Second, the implementation of manufacturing-aware optimization incorporates process tolerance constraints and geometric feasibility checks, effectively bridging the gap between simulation-based optimization and physical fabrication requirements. Third, comprehensive algorithm benchmarking reveals that the modified Nelder–Mead method achieves a superior convergence efficiency (~100× computational efficiency) compared with evolutionary algorithms, while maintaining design quality, providing critical advantages for rapid MEMS prototyping. The platform establishes a new paradigm for MEMS co-design through tight integration of multiphysics simulation, manufacturing constraints, and intelligent optimization algorithms.

Driven by “More than Moore”, miniaturization and multifunctional integration of micro-energy devices are emerging as critical pathways for next-generation compact microsystems. This study proposes a sensing-in-Energy (SiE) microdevice that immerses an inertial switch in a parallel-connected supercapacitor’s electrolyte, enabling simultaneous impact sensing and stable energy supply under extremely high gravitational acceleration (high-g) shocks (over 10,000 g). The SiE microdevice can be viewed as a high-amplitude shock sensor (raw signal peak > 50 mV) under high-frequency perspective, and a shock-resistant electrochemical power source (voltage fluctuation < 2%) under low-frequency perspective, while energy consumption reduces over 99.9% compared with conventional high-g sensor due to its event-driven mechanism. Sensing performance is boosted > 50% using multiphysics model combined with machine learning algorithm. Furthermore, a fuze microsystem was built based on SiE microdevice, achieving 150 μs-level ultrafast response. Three-layer penetration experiments have verified the engineering application of SiE microdevice and its fuze microsystem in smart munitions domains, providing a novel paradigm for heterogeneous microsystem in high-dynamic environments.

Parasitic capacitance is an unavoidable and usually unwanted capacitance that exists in electric circuits, and it is the most important second-order non-ideal effect that must be considered while designing a triboelectric nanogenerator (TENG) because its magnitude is comparable to the magnitude of the TENG capacitance. This paper investigates the structure and performance optimization of TENGs through modeling and simulation, taking the parasitic capacitance into account. Parasitic capacitance is generally found to cause severe performance degradation in TENGs, and its effects on the optimum matching resistance, maximum output power, and structural figures-of-merit (FOMs) of TENGs are thoroughly investigated and discussed. Optimum values of important structural parameters such as the gap and electrode length are determined for the different working modes of TENGs, systematically demonstrating how these optimum structural parameters change as functions of the parasitic capacitance. Additionally, it is demonstrated that the parasitic capacitance can improve the height tolerance of the metal freestanding-mode TENGs. This work provides a theoretical foundation for the structure and performance optimization of TENGs for practical applications and promotes the development of mechanical energy-harvesting techniques.