Perovskite solar cells (PSCs) exhibit a well-documented hysteresis in their current-voltage (I-V) characteristics, the origin of which has been rationalized in terms of surface boundary-induced capacitance effects [1]. This hysteresis renders conventional I-V sweep-based power conversion efficiency measurements unreliable, and maximum power point tracking (MPPT) has been established as a more rigorous approach to PSC characterization at the laboratory scale [2]. Translating this methodology to field deployment and production environments, however, introduces new challenges: dynamic and non-uniform illumination, long-term environmental degradation, and the need for rapid characterization compatible with manufacturing throughput. This work presents three embedded MPPT-based systems developed to address these challenges in the context of real-world PSC deployment.

Neural Network MPPT for Partial Shading Conditions: Conventional hill-climbing MPPT algorithms are susceptible to convergence at local power maxima under partial shading — a condition frequently encountered in vehicle-mounted or urban-deployed PSC panels where illumination is spatially and temporally non-uniform. A custom-trained neural network, implemented on an ARM Cortex-M4 based microcontroller, was developed to identify the global MPP under such conditions. A 3-month outdoor comparative study using two arrays of four PSC panels each, subjected to controlled dynamic shading via motorized wipers, demonstrated a mean power extraction advantage of 5.9% over a hill-climbing controller, with a peak advantage of 17.4% observed under shading geometries inducing local maximum trapping. No advantage was observed under uniform full illumination, confirming the result is attributable to superior global MPP identification rather than tracking speed alone.

Self-Updating Dual Lookup Table MPPT for Degradation Monitoring: A second strategy employs two lookup tables maintained on a microcontroller, each storing Vmpp values as a function of real-time temperature and irradiance. One table retains the initial baseline Vmpp map; the second is continuously updated with current values. Ongoing comparison between the two tables allows environmentally resolved tracking of PSC performance drift — distinguishing reversible environmental dependence from irreversible degradation under specific temperature and irradiance conditions. In an 8-week outdoor trial, a Vmpp shift of approximately 1% was detected and correlated with environmental parameters, demonstrating the system's sensitivity to early-stage PSC degradation without requiring laboratory instrumentation.

Photo-Triggered Flash MPPT Tester: A third system, currently under development, targets rapid MPPT-based PSC characterization within a single light pulse, which will be discussed in this presentation. Together, these systems extend the MPPT-based characterization paradigm from the laboratory to field deployment and manufacturing contexts, providing new tools for investigating PSC stability and performance under real-world operating conditions.