Purpose <p>The high cost of commercial multispectral monitoring systems has limited widespread adoption in agricultural research and production systems.</p> Aim <p>The study aimed to design and develop a low-cost multispectral monitoring system (I-See-18B), validate the responsiveness of the I-See-18B with a commercial reference device and evaluate the performance of the I-See-18B under low and high solar irradiance conditions to determine the operational limits of the system.</p> Methods <p>The I-see-18B system was assembled using open-source, Arduino-compatible electronic components at an approximate cost of US$250 (parts only). The sensing unit comprised upward and downward SparkFun Triad Spectroscopy sensors. The I-see-18B system was calibrated using MAPIR grayscale calibration targets and low-cost BEHR gray and green scale color chips. Field trials were conducted on different surface types during spring 2023 (low irradiance) and summer 2025 (high irradiance) to evaluate system performance. NDVI values derived from the I-see-18B were compared with a commercial reference sensor (GreenSeeker) using regression analysis.</p> Results <p>NDVI values measured in agricultural fields using the I-See-18B exhibited a strong correlation with GreenSeeker measurements (R² = 0.953), indicating reliable performance for field-based vegetation monitoring. The upward sensor retained accuracy under low irradiance but exhibited saturation under high irradiance, whereas the downward sensor consistently provided stable reflectance data in both high and low radiance. A spectral band normalization approach effectively reduced irradiance effects, indicating that the upward sensor could be omitted and further reduce costs.</p> Conclusions <p>Field evaluation demonstrated that the low-cost I-See-18B system can support site-specific agricultural applications for assessing soil and crop variability, within the limitations associated with illumination variability and sensor response behavior.</p>