wirSENSE: A General WirelessHART Sensor Platform
Master thesis
Permanent lenke
http://hdl.handle.net/11250/259547Utgivelsesdato
2008Metadata
Vis full innførselSamlinger
Sammendrag
The development of wireless sensor networks (WSNs) has grown rapidly during the last few years. Much because of the introduction of the IEEE 802.15.4 standard in 2003, which provides the basis for several WSNs. The ZigBee standard, released in 2004, was the first open and complete WSN standard based on the IEEE 802.15.4. Unfortunately, previous work has shown that ZigBee is not well suited for industrial applications. The lack of an open standard covering all layers, has led to the development of several proprietary solutions. Large companies like StatoilHydro are cautious to go for a proprietary solution because of the dependency to a single supplier and therefore, they are not using WSNs in a large scale until an approved standard has emerged. WirelessHART is a new open standard for wireless sensor networks designed for industrial use. This standard is based on the lower layers of the IEEE 802.15.4 standard and the higher layers of the traditional HART standard. HART is much used in industrial instrumentation and Dust Networks is the first company that claims to have a WirelessHART compatible system with their new SmartMesh IA-510. At the time of writing this report, the system is yet to be released, but some early samples of the SmartMesh IA-510 system were available for use in the thesis work. The WiVib is a wireless vibration monitoring system developed by ABB, SKF, Sintef and StatoilHydro. It is currently a prototype and it is using an earlier non-WirelessHART compatible version of the SmartMesh system. The final version of the WiVib is supposed to be WirelessHART compatible. This system has been studied and tested. The main focus of this master's thesis has been to develop a sensor platform based on WirelessHART compatible devices from Dust Networks. The platform is supporting several different sensor types and easy integration of new sensor types is important. In addition, it is small in size and battery driven. The existing WiVib prototype is used as inspiration, mainly because of the radio system. WiVib is also tested to verify the performance of the radio. The results indicate that the lower layers of the WirelessHART protocol are capable of handling the dificult industrial environments at StatoilHydro's industrial lab at Rotvoll. It was possible to run a 12 hour test with full mesh topology, keeping reliability at 100.00% and the latency below 3 seconds. The maximum range of the sensor nodes is approximately 50 meters. Test shows that the WiVib system has problems handling WLAN noise over longer time periods. In one case, reliability decreased to 92.60% and latency increased to 9.5 seconds. The wirSENSE design is proposed as a general sensor platform design. This is a design where the sensor circuit is separated from sensor independent hardware. The sensor interface is common for all sensor types. Four wirSENSE sensor node prototypes are built to prove the wirSENSE design. In addition, a wirSENSE GUI and Configuration Tool is made. The prototypes are tested in a similar, but less extensive way than the WiVib nodes. The best antenna had 50 meter range, network behavior was in line with the WiVib performance and battery consumption was lower than on the WiVib (79µA). The conclusion is that the wirSENSE sensor mote design is a success. The wirSENSE GUI and Configuration Tool is the PC application that is controlling the wirSENSE sensor motes. The GUI is written in Ruby and communicates with a network manager. Information is exchanged using an XML interface defined by Dust Networks. Although there are some minor problems with the GUI, it serves its purpose as a demonstration tool for the wirSENSE system. Unfortunately, the HART message format is not used by wirSENSE motes and GUI. This is due to the lack of high level HART support in IA-510 and because of time constraints. It was considered more useful to build a working prototype, and demonstrate the possibilities of IA-510, than to implement all the HART message types in software. It is concluded, that Dust Networks should improve the WirelessHART support, to ease development of new WirelessHART products.