It’s a challenging problem to track the location of people and assets indoors, on multiple floors, in a multi-building campus. It’s even more challenging to solve this problem with a technology that also meets the business needs of cost, accuracy, and ease of installation. The client gave us thirty days to create a Proof-of-Concept to demonstrate a solution that met these disparate goals. After we successfully demonstrated the underlying technology, we began development on a limited-coverage-area system, which was then successfully field-tested. The next stage of the project was to increase the system’s capabilities to track thousands of assets across a large, corporate campus environment. This project demonstrated both our broad technical capabilities (system engineering, electrical design, software development, mechanical engineering,and integration & test) as well as our ability to lead and manage a complex system through the various product-development phases.
Understandably, a system this complex requires a number of trade-offs during the design process. Our job was to optimize the ones that most affected the system’s cost, complexity, and performance. Early key decisions involved the choice of wireless technology, and choice of power source (line vs. battery) for the location-centric infrastructure assets that needed to be deployed on a wide scale – one per room. After analysis and discussions, we choose the 802.15.4 wireless standard (the standard that ZigBee is built on). Use of 802.15.4 transceivers allowed battery operation of the location-centric infrastructure assets, with a battery life of greater than five years. The choice of battery power also enabled them to be installed in literally less than a minute. This technology choice provided the client a significantly lower installation and support cost structure than any competing solution. Our ability to see, understand and take advantage of these types of key technology decisions and architectural choices provides you the innovative solutions you are looking for.
Once the key technology decisions were made, development began in earnest on a limited coverage system. In this phase, the key focus was integration and test. The various elements of the system were first tested individually. Then the hard part begins, making all the elements work together – the integration and test phase. System integration and test is the portion of a project that is historically underestimated, and often represents the majority of the effort. The first objective during this time is to get the basic system functional in the lab. Once that goal is achieved, additional features may be enabled and more complex operational scenarios tested. Only after a system passes lab testing, is it allowed to proceed to field trials. Throughout this process, our development team finds and fixes software bugs, improves the hardware and refines our analytical models of the system to optimize system performance. In the case of the Indoor RF Location system, the extensive testing we had done before the field trial laid the ground work for a successful field deployment.
The successful field trial meant the next phase was about one objective: scale. The system had to be able to track thousands of assets, and provide both the current location and a location history of each asset. Further, the system’s infrastructure now was spread across a large corporate campus. Lastly, this iteration of the project allowed multiple applications to access and consume the location data. This phase of implementation and deployment truly tested the initial design decisions we made early in the project, demonstrating that each of the choices we had made allowed for the extensibility to scale the system as needed, without compromising the basic criteria of cost, accuracy and ease of installation. In the process of increasing scale, we enhanced each element of the system, improving manufacturability of all hardware elements, increasing robustness of embedded software, and significantly augmenting the capacity of the server software. Besides redesigning and reworking the various elements to improve their performance, we also greatly expanded the test suite. We built simulators and other test support equipment to stress test the system at twice its advertised level. This margin provides insurance against the variability of real-world operation and the criteria that the system meets in the test lab. As more was learned about system behavior, the test suite was expanded to include those situations.
Successfully executing a project of this scale required us to exercise our full arsenal of product development capabilities. These are the types of opportunities at which we excel.