Intelligent Networked Systems Lab


The next generation networked and data-driven systems will be all about “intelligent” machines where machines will learn how to function when exposed to varying conditions surrounding them. The INS lab at UMKC envisions developing computing and communication techniques for building intelligent networked systems.

Current Students:

PhD

Masters

K-12

Edge intelligence for Data-driven Systems: In the era of Big Data, extracting values from data is of great importance for wellness, safety, and other social good. This value proposition, in part, is due to the proliferation of heavily instrumented physical spaces with a rich and diversified set of sensors constituting the so-called the Internet of Things (IoT). Pushing all these data to the cloud for processing is neither viable nor economical; instead, recent efforts suggested an edge-centric approach that brings the computation closer to data and users thus giving rise to an edge computing paradigm for Big Data (intelligence at the edge). Future data systems, therefore, will make extensive use of data analytics/machine learning that would enable data-driven decisions. As a result, computing tasks pertaining to training a model or using a trained model will constitute a large part of computation at the edge. Therefore, in addition to designing and training these models (where most efforts concentrate these days), novel efforts should be undertaken to manage and deploy these models, especially when they are deployed at the edge. At the same time, the edge itself will manifest new challenges and opportunities in designing new ML algorithms tailored for power- and compute-limited edge devices.

Data Systems for Societal CPS/IoT: Societal CPS assumes humans are the explicit participants of the physical world. The cyber part, therefore, should consider humans as active agents that generate information (sense), process information (compute), remember information (store), pass information (communicate) and take actions (actuate). The presence of human as part of a computing system brings non-trivial complexities in terms of modeling their behavior and role, which calls for research challenges in developing novel computing theories in societal contexts. Future societal IoT/CPS applications and Big Data systems at large will leverage techniques from both network and data science—a synergic integration of information networks, social networks, communication networks, and cognitive networks.