Information Density as a Quantitative Measure for AI-enabled Virtual Sensing: Feasibility and Limits

Researchers propose Information Density as a quantitative framework for optimizing IoT sensor networks by enabling virtual sensing through AI. Using spatial, temporal, and cross-modal correlations, the system can replace physical sensors with computational models while maintaining sub-4% error margins, demonstrated via Madrid's smart city infrastructure.

The paper addresses a critical infrastructure challenge: IoT networks generate exponential data volumes that strain storage, bandwidth, and processing capabilities. Traditional compression techniques sacrifice data integrity or computational efficiency. This research pivots toward a fundamental question—whether physical sensors are necessary at all if spatial and temporal relationships between data streams contain sufficient information to predict unmeasured variables. The proposed framework quantifies this potential through two complementary metrics: Phase in Eigen Space and Mutual Information, both mathematically grounded in information theory. The approach leverages correlations across both single-modality sensors (e.g., multiple temperature readings) and cross-modality combinations (e.g., temperature predicting humidity). Real-world validation using Madrid's smart city data demonstrates practical feasibility, achieving sub-3.21% mean error with minimal sensor coverage. This has profound implications for IoT deployment economics. Smart city operators, industrial facilities, and environmental monitoring networks could substantially reduce hardware costs, power consumption, and maintenance burdens by replacing redundant physical sensors with virtual ones. The energy savings alone could accelerate adoption of comprehensive monitoring in resource-constrained regions. However, the framework's effectiveness depends on pre-existing data correlation patterns; scenarios with unpredictable or truly independent variables may not benefit. Future work should explore scalability across heterogeneous sensor types, robustness to sensor failures or anomalies, and applicability beyond stationary urban infrastructure to dynamic environments. The intersection of information theory and edge AI here signals a maturing field where theoretical optimization directly enables practical infrastructure efficiency.

• →Information Density quantifies whether virtual sensors can replace physical ones using AI-driven correlation analysis across sensor networks.
• →Framework achieved sub-3.21% mean error in Madrid smart city tests, validating feasibility of sensor reduction at scale.
• →Two complementary metrics—Phase in Eigen Space and Mutual Information—enable optimal sensor placement and virtual sensing configuration.
• →Reduces IoT infrastructure costs, power consumption, and maintenance while maintaining measurement accuracy in smart environments.
• →Effectiveness depends on spatial and temporal correlations in data; independent variables remain unmeasurable through virtual sensing alone.