Manufacturing processes would greatly benefit from fusing data from many disparate sensors, but systems today do not fully exploit available sensor data. Disparate sensors could include Coordinate Measurement Machines (CMM), laser surface scanners, micro sensors, cameras, acoustic devices, thermocouples, or other various devices which provide measurement or visual data. Often, sensor data requires separate customized software for each type of sensor system, as opposed to having common tools for use across a wide array of sensor systems. This process of stove-piping requires proprietary software for analysis and display of each sensor type, and inhibits interoperability. There are several challenges to sensor fusion which need to be addressed. First, many sensors providing data are heterogeneous in phenomena detection and operation, providing measurements of different target attributes. This makes the measurements very difficult to fuse directly. Second, these disparate sensors are asynchronous in time. The collection, integration, buffering and transmitting time can each affect the way time is calculated and stored by the sensor. Transducer Markup Language (TML), developed by IRIS Corporation, addresses these challenges. This paper describes TML and addresses examples of industrial applications of TML-enabled transducer networks.

This content is only available via PDF.
You do not currently have access to this content.