As sensor technology increasingly pervades the physical world, computer systems are monitor and control the physical world in a new way that could revolutionize many areas of science and engineering. Wireless, battery-powered devices offer the promise of anywhere, anytime information that could improve transportation systems, factories, buildings, and other aspects of the human environment. However, these sensing systems are often too complex for non-specialists to design, build, and deploy and software is one of the main sources of complexity: ubiquitous sensing systems combine aspects of the most challenging computer systems, including distributed systems, embedded systems, wireless systems, and low-power systems.

The aim of this project was to develop new technology to easily manage this complexity, enabling scientists and engineers to use wireless sensing systems just like other tools and instruments. Specifically, this research has taken a comprehensive approach to a new technology called “macroprogramming” -- the task of programming an entire network of devices as a single, programmable substrate.

During the project, we created new programming abstractions that allow the separation of the application logic from quality-of-service requirements and hardware requirements. We applied these techniques first to stationary sensing networks and demonstrated that the same code could be executed efficiently on networks of many different sizes and topologies without modification. Thus, this work improves code portability and reuse for wireless embedded systems. We then extended these techniques to mobile sensing networks, also called “drone sensor networks”, and showed that the same code could execute efficiently even as the network dynamically changed shape, size, and organization. Again, this result demonstrates improved code portability and reuse than state of the art techniques.

In addition to improved programming abstractions, we developed new techniques that use global information a macroprogram to optimize system performance, automatically generate test cases, and reduce the state space for analysis. This research exploits global, network-wide information about a sensing system provided by a macroprogram to improve traditional software engineering techniques such as testing, debugging, analysis, and optimization. The artifacts of this research was combined into a comprehensive development environment for CPSs called MacroLab. The new tools developed will greatly simplify the process of programming both stationary and mobile sensing systems and make them more accessible to non-experts. By taking a holistic view of the network and its software, MacroLab can manage a range of complex, interacting issues that would be extremely difficult to manage by hand.

This project resulted in over a dozen peer-reviewed publications and over one-hundred fifty citations, supported the graduate programs of three PhD students, and supported publications from two undergraduate researchers, and supported one international collaboration. The research projects are documented online and the publications have been made available to the general public and to other researchers. 

Last Modified: 07/18/2015
Modified by: Cameron (Kamin) D Whitehouse