Technical Limitations and Research Gaps of IoT and Big Data Infrastructures in Precision Crop Production: A Design-Oriented Review

Precision crop production increasingly relies on Internet of Things (IoT) devices, heterogeneous sensor networks, machine telemetry, and data-intensive analytics to monitor field conditions, support decisions, and enable variable-rate or autonomous operations. However, farm-scale multi-season adoption remains limited by technical constraints that are often reported only as general challenges. This design-oriented review clarifies those constraints at the infrastructure level. It is not a quantitative meta-analysis; rather, it combines a transparent multi-database search, a PRISMA-type selection record and thematic design synthesis of the 2020–2025 literature, supplemented by selected 2026 studies and current interoperability and security specifications. The review addresses four research questions covering field hardware and connectivity failures, edge-to-cloud data management, integration with farm management information systems and agricultural machinery, and future design priorities. The synthesis identifies recurring gaps in multi-season reliability evidence, calibration and self-diagnostics, energy and connectivity benchmarking, operational definitions of agricultural Big Data, metadata/FAIR implementation, ISO 11783/ISOBUS–FMIS interoperability, lightweight cybersecurity, and serviceability. The main outputs are an evidence-traceability matrix, a distributed reference architecture specifying inputs, outputs, standards, validation points and edge/cloud placement, an operationalized G1–G10 gap matrix with indicators and evaluation designs, and minimum reporting requirements for future agricultural IoT studies. These outputs are intended to make field systems more interoperable, maintainable, secure, and evaluable. Because the corpus combines heterogeneous evidence types and does not support quantitative meta-analysis, the outputs should be interpreted as design and reporting guidance rather than comparative performance estimates.