Robotics in agriculture and forestry

Robotics in agriculture and forestry

In agriculture and forestry, robotics has made a substantial impact. Farmers are conscious of their need for automatic vehicle guidance to minimize damage to the growing zone of their soil. Automatic sensing, handling, and processing of produce are now commonplace, while there is substantial instrumentation and mechanization of livestock procedures. In forestry, legged harvesters have not yet seen great success in their application, but the automation of trimming and forwarding with simultaneous localization and mapping techniques will change the industry in the future.

Some impressive developments in walking forestry harvesters are presented, including machines targeted towards the difficult terrain of the Scandinavian forests. More-conventional cut-to-length harvesters are also highly automated, while operations such as delimbing must be carried out at speed. Before complete autonomous harvesting becomes possible, some of the localization and mapping techniques that are described must come to fruition.

The combination of machine vision with global positioning by satellite (GPS) allows a tractor to follow a row of crops, performing a headland turn at the end of the row. The history of a series of projects is outlined, leading to the present outcome that is in the process of being commercialized. Another project that is based on machine vision relates to the location of macadamia nuts. To select which trees should be propagated, it is necessary to attribute fallen nuts to the correct tree. Color sorting and grading of produce is not a matter of sensing alone, but involves a measure of produce handling that puts it at the fringe of robotics.

Automated milking parlours have proved their worth. However success has eluded some other projects described here, such as automated sheep-shearing and an automated abattoir. Another project is presented that literally sorts the sheep from the goats, using a swinging gate to separate different species using machine vision so that feral species are excluded from watering holes in the dry Australian outback.

Although robotics is making rapid inroads into these areas, they are still a fruitful source of application projects, some sufficiently demanding to require the development of new theoretical techniques.

© Springer 2008.