GT IoT Projects

GT IoT Projects
Fumin Zhang
Associate Professor, Systems and Controls, School of Electrical and Computer Engineering

The Principal Investigator (PI) creates a research program to establish a theoretical foundation for analysis and design of control tasks on networked embedded computing systems powered by battery. In modern day applications, feedback control laws are typically implemented as control tasks that are scheduled and executed in real time networked embedded computers under constraints imposed by limited energy budget. The PI establishes a new approach to manage computing delay and sampling period associated with real time scheduling as feedback control mechanism. The goal is to find better ways to integrate modern control methods such as the model predictive control with the new generation of real-time systems. Since mobile devices supported by batteries have become pervasive, battery related failures have raised public attentions on the safety and reliability of high tech equipment. The PI develops novel methods for battery failure detection that may be applicable to large packs of batteries. The method proposed is based on an adaptive filter that has strong connections with the universal adaptive stabilization method.

Dr. Zhang is the Interim editor in chief of "Cyber Physical Systems", a new international interdisciplinary journal dedicated to publishing the highest quality research in the rapidly-growing field of cyber-physical systems and the Internet of Things (Taylor & Francis).

Andrew Dugenske
Director, Factory Information Systems Laboratory

The objective of this project was for Georgia Tech to assist its partner in the design of a mechanical pump, fabrication of electronic controls, and production of monitoring software to maintain landfill leachate levels with the goal of optimizing decomposition and maximizing landfill gas to generate electricity. In order to reduce cost and increase modularity, a prototype control system based upon a single board computer (Beagle Bone Black) was produced. By making use of commercial off the shelf (COTS) parts that are well supported, easily expanded and simply programmed, the Georgia Tech partner can now rapidly enhance the pump with customer-desired features.

Georgia Tech also developed a prototype cloud application that is capable of receiving data from pumps and display the data in charts and tables. The cellular communication and cloud application will allow the partner to offer pump monitoring services that has the potential to greatly reduce landfill operator costs by eliminating the need for company personnel to visit pumps on a time based maintenance schedule. Rather, the appropriate resources can be dispatched to the appropriate pump when the data from the pump indicates maintenance is required.

This project helped establish a high growth manufacturing company in Georgia by accelerating a new product and services to market. The product being produced by the GT partner is also facilitating energy production from a domestic ‘green’ source. Project funding came from various sources, i.e.,  the partner, the Georgia Centers for Innovation (COI), and the Georgia Manufacturing Extension Partnership (GaMEP). In addition, many undergraduate and graduate students contributed to and benefited from the project. An ME (Mechanical Engineering) 8843 (advanced mechatronics) team of three graduate students devoted an entire (spring 2014) semester to the project and presented their work for a class grade. One of the students continued working on the project as a special topics course during the summer of 2014.Two students from the Georgia Tech Manufacturing Institute (GTMI) hosted Research Experience for Undergraduates (REU) program worked exclusively on the project over the summer of 2014 and three visiting students from Tunghai University (China) contributed to the project during their four week visit.

Andrew Dugenske
Director, Factory Information Systems Laboratory

This project dealt with a wide range of parameter monitoring, i.e., wildlife control, temperature, electricity, natural gas, water, steam and equipment status. Typically, this type of Internet-of-Things (IoT) solution consists of electronic devices in the field that gather data and transmit it to a cloud application so the data can be viewed, alerts communicated to the appropriate individuals, and messages sent to machines for further action. The primary purpose of this project was to develop a new cloud application that allowed Georgia Tech’s partner to address a variety of industries, customers and devices.

A comprehensive cloud-based software application was developed that can assist the partner to serve multiple industries.  In order to build an application that is flexible, scalable, maintainable and easily expandable, a layered architecture was used among the major components. Interfaces among the layers were precisely defined, so that any of the components can easily be swapped-out with an alternative if the need arises. A software architecture of this type provides a more formalized method of constructing an application so errors and performance parameters can be measured among the various interfaces. A layered approach also allows developers to focus on their particular area of expertise (i.e. database, server programming, user interface) because the interfaces provide them with a clear line of demarcation and functionality.

Various technologies were used to implement the application. The Georgia Tech hosting service was the primary platform used to host the development, which is a Parallels virtualized environment running Red Hat enterprise Linux. The application and web server are Apache 2.2 that runs PHP scripts to accomplish server-side scripting. The database is MySQL and significant effort went into developing the stored procedures that provided functionality for the PHP code. Drupal 7 is used as the content management system. JavaScript based upon the YUI framework is embedded into Drupal pages to provide the user with a dynamic experience. The exchange of information between the JavaScript and the database was accomplished using XML over HTTP. Bit Bucket was used as the distributed revision control system for all of the source code.

More information can be obtained from Georgia Tech’s Factory Information Systems (FIS) Laboratory (http://fis.gatech.edu/)

Lee W. Lerner
Research Faculty, Georgia Tech Research Institute

Our overarching research goal is platform-based architectural support for, and rigorous validation of, Trust Enhancement of Critical Embedded Processes (TECEP). We are guided by a design philosophy that the most trusted layers of a system should validate requests from less trusted layers, and otherwise take corrective actions. While this may not be easily accomplished in systems interacting with people, process control systems (PCSes) have the advantage of precise specifications and accurate models that can be used to check the immediate or eventual consequences of a controller’s malicious or inadvertent actions.  For PCSes we are developing a Trustworthy Autonomic Interface Guardian Architecture (TAIGA) that provides an on-chip, digital, security version of classic mechanical interlocks. In order to enhance trust in critical embedded processes, TAIGA redistributes responsibilities and authorities between a general purpose processor and a hardware-implemented interface controller, simplifying controller software without significantly degrading performance while separating trusted components from updatable code. The interface controller is synthesized from high-level source code, formally analyzed, and permits runtime checked, authenticated updates to certain system parameters but not code. TAIGA’s main focus is ensuring process stability even if this requires overriding commands from the processor or supervisory nodes. The TAIGA architecture is mapped to a commercial, configurable system-on-chip platform, which can optionally host the PCS controller.

Pages