Research & Publications

The research programme at SIḌḌANTA — Systemic Intelligent Data-Driven and Network-Theoretic Analysis — is built on a conviction that genuinely useful intelligence requires more than data and computation. It requires systems-theoretic rigour, domain knowledge, and an understanding of how variables causally relate across networks.

Three foundational ingredients — multi-modal Data, Systems Theory, and Domain Knowledge — converge through AI & ML and Network Science to produce actionable capabilities: learning dynamic models, forecasting under uncertainty, monitoring and diagnosing faults, performing causal analytics, and designing intelligent sensing systems.

Four broad research directions organise the lab's active work, each addressing a different face of the fundamental question: how do we extract reliable, interpretable knowledge from complex, multi-modal data streams?

These directions are not silos — a project on federated learning for distributed sensors also demands causal discovery methods, and work on intelligent transportation systems draws simultaneously on sensing, analytics, and control. The connections are as important as the directions themselves.

The lab's research spans three decades and two institutions — University of Alberta, IIT Madras, and now IIT Tirupati. What began as multirate control and multiscale process monitoring has grown into a broad programme connecting methods research with real-world applications across process engineering, climate, transportation, agriculture, seismic analysis, and structural systems.

The timeline below shows when each thread was initiated and its current status. Methods threads (top half) develop the theoretical and algorithmic foundations; application threads (bottom half) drive the problems and validate the methods. Most active threads date from 2010 onwards, reflecting the shift towards data-rich environments and AI-driven discovery.