Sensor Networks in Agriculture
Wireless sensor networks can provide farmers with real-time measurements of parameters such as soil moisture, crop health, and animal movement. This data will help inform practices for sustainable agriculture and help farmers to more accurately and effectively control activities such as irrigation, planting, stock movement, and pesticide application. Networks of nodes with both sensors and actuators will eventually not only monitor the agricultural environment but also control it intelligently.
Virtual Fencing
CSIRO's Virtual Fencing technology involves smart collars worn by cattle which allow monitoring of their location and movements. These collars can be used to restrict the movement of cattle past "virtual fences" in paddocks or around environmentally sensitive areas.
The collars contain a wireless sensor node (based on CSIRO's Fleck™ technology) and other sensors for inertial measurement and global positioning (GPS), as well as a power supply, radio transceiver and low power microcontroller which stores the GPS data for the virtual fencelines. If a cow approaches a virtual fence the collar emits a warning sound, and if it continues past the fence line the cow receives a tactile stimulus*. In trials cattle quickly learn to react to the audio warning and stop before the virtual fence.
The paddock also contains static wireless sensor nodes for environmental monitoring, and these static nodes exchange data with the mobile sensors nodes on the cattle. The mobile sensors upload their GPS locational and inertial postural information to the static nodes for collection and analysis, while the static nodes can be used to distribute updated boundary fenceline information to the mobile nodes.
As well as providing a system for containing cattle, the technology also allows for a study of animal behaviour as conditions change in the paddock. Research in this area brings together ICT researchers with scientists from CSIRO's Livestock Industries Division. It is hoped that a better understanding of the dynamics of herd hierarchy will allow herds to be controlled by only collaring some of the animals.
*The tactile stimulus is much milder than the shock experienced by an animal which touches an electric fence, and animal ethical standards are met in all our experiments.
CSIRO's Virtual Fencing project was the Queensland State Winner of the Research and Development category in the Australian Information Industry Association's iAwards which recognise outstanding ICT innovators.
View the other 2009 iAwards winners here.
Bull Separation
During breeding season, bulls in the same paddock can act aggressively towards each other and may inflict serious injuries. CSIRO researchers have developed a way to keep bulls separated through the use of smart collars.
Each bull wears a collar which contains a wireless sensor node, a global positioning (GPS) sensor, rechargeable batteries, a radio transceiver, a data-logger and a harness for applying an electrical stimuli to a bull.* The collar transmits the location and velocity of its host bull to the collars on all the other bulls, forming an ad-hoc network of mobile nodes. Analysis of the location, velocity and the direction the bull is heading are performed in real-time to assess the "state" of a moving animal. If the bull is determined to be acting aggressively towards another bull, as indicated by its velocity and heading, then a mild electrical stimulus is administered to curb the bull's aggressive behaviour. In this arrangement, there is no coordination via a central node - all communication is peer to peer between the smart collars, and all processing is performed "on-board" the bull. Field trials have proven the technology successful in autonomously separating aggressive bulls.
Future research will investigate the effectiveness of the autonomous bull separation technology in larger paddocks where direct transmission to all nodes is not always possible due to the large distances involved. This scenario will require multi-hop communications to ensure that information is shared between all network nodes.
*The collars satisfy all animal ethics requirements.
Pasture Management
A wireless sensor network (WSN) of mobile and static nodes deployed on pastoral land is creating the "smart farm" of the future. The sensor network helps to prevent overgrazing or undergrazing of the land, and prevents grazing animals from accessing environmentally sensitive areas. The system simultaneously attempts to understand the state of the climate, soil, pasture and animal movements.
Static Fleck™ wireless sensor nodes are deployed around the paddock, with each node containing an onboard temperature sensor, radio transceiver, solar rechargeable batteries and either a soil moisture sensor or a CMOS camera module and Flash card for imaging. The soil moisture sensors detect the volumetric water content of the soil and transmit this information back via the network of nodes to a base station for aggregation. The CMOS cameras take regular images of the pasture, and the compressed images are collected at the base station for analysis of grass coverage, height, and greenness.
Mobile Fleck™ wireless sensor/actuator collars placed around the necks of cattle are used for the management of cattle. (See Virtual Fencing above). The nodes monitor the position of the animals and the in-built actuators give auditory and electrical warnings to the cattle to prevent them from breaching virtual fencelines.
Contacts
Publications
-
Ying Guo, Geoff Poulton, Peter Corke, Greg Bishop-Hurley, Tim Wark, and David Swain. "Using accelerometer, high sample rate GPS and magnetometer data to develop a cattle movement and behaviour model". Ecological Modelling, Volume 220, Issue 17, 10 September 2009, Pages 2068-2075.
- Tim Wark, Chris Crossman, Wen Hu, Ying Guo, Philip Valencia, Pavan Sikka, Peter Corke, Caroline Lee, John Henshall, Julian O'Grady, Matt Reed and Andrew Fisher. "The Design and Evaluation of a Mobile Sensor/Actuator Network for Autonomous Animal Control". Proceedings of Information Processing in Sensor Networks (IPSN 2007/SPOTS 2007), Cambridge (MIT Campus), Massachusetts, USA, April 2007.
