A model for assessing reusable, remoting sensors in test and measurement systems

In a typical system, sensors communicate with a computer via a communication port, such as a serial port; however, with the recent advances in programming languages and the addition of many low-cost networking devices, new methods for communicating with sensors are essential.

A sensor's communication port provides an interface for connecting the sensor with a computer to control the sensor or acquire data. Most sensors utilize the Recommended Standard 232 (RS232) port for communication with a computer; however, this port was created to serve as an interface for computers and modems to communicate, not sensors. Utilizing the RS232 port to communicate with a sensor requires configuring the computer's and sensor's settings, cables, and commands. As a result, a majority of sensors are not "plug and play" and additional time is required to configure and develop the software to communicate with the sensor. The use of models, modern computers, and networking solutions can be utilized to alleviate these problems and improve the way sensors are incorporated into the design of a system.

Therefore, this thesis creates a model for the development of reusable remote servers which handle the communication interface to a sensor and provide a network interface to a computer to control and acquire data from the sensor. The model created can be modified to support various communication interfaces a sensor may use to communicate with a computer.

The model is designed with the Unified Modeling Language (UML) to provide reusable diagrams of the model for the development of a sensor subsystem. The diagrams are applied to several sensors by using development platforms and languages to create a sensor’s communication software. The sensor's communication software is deployed on a network processor and the sensor is connected to the network processor which results in the creation of a sensor subsystem. For each sensor subsystem, a software executable is deployed on a networked processor that communicates with the sensor and allows a client application on a different network processor to connect to the sensor subsystem and interact with the sensor.

The sensor subsystems created through the diagrams include a heise pressure subsystem, a vaisala temperature/humidity subsystem, and a mettler mass comparator subsystem. The time required to create the sensor subsystems through the diagrams is measured and recorded to determine if the diagrams created can be reused to reduce the amount of time required to interface and communicate with a sensor through a computer. In addition tests are conducted to determine if any advantages can be found by utilizing sensor subsystems in the creation of test and measurement systems. For example, development time of the client application is measured to determine if the use of sensor subsystems can reduce the amount of time required to develop a complete test and measurement system. Additional tests are performed to explore other advantages the sensor subsystems may provide and the results of the tests are compared with prior research.

Finally, the results of the tests conducted utilizing the sensor subsystems to develop test and measurement systems have shown reduction of development time of client applications by 60%. In addition, the use of the sensor subsystems also enable multiple client applications the ability to share the sensors allowing the sensor to be reused in the design and deployment of different test and measurement systems. The results of the tests conducted have shown that utilizing UML modeling tools to create diagrams for developing sensor subsystems is an effective means of reducing development time of a sensor's communication software by 90% or more, and improves sensor configuration. Furthermore, the use of sensor subsystems enables reuse of a sensor in other systems.