KUCHING: The agriculture sector should make use of digital technologies such as robotics and machine vision to control weeds, says Sarawak Deputy Chief Minister Datuk Amar Douglas Uggah.
He said there was a need to develop new tools and techniques for effective weed management in view of the manpower shortage in agriculture.
“This is especially important in large and commercial-scale farms. Technologies are already available on the application of robotics and machine vision for weed detection and removal.
“These digital technologies could revolutionise weed management leading to the year 2050,” he said when opening the 27th Asian-Pacific Weed Science Society (APWSS) conference here.
Giving an example, Uggah said hyperspectral imaging methods could now be used to detect weeds, regardless of the plant’s visibility or distinct leaf shape.
“Studies show that it can reliably distinguish closely-related plants, for example tomato and black nightshade,” he said.
Another promising technique, Uggah said, was the systems approach to develop smart machines for automated weed control.
In this technique, crop planting locations are mapped and a weeding robot can then access the map to distinguish the crops from weeds based on location.
Uggah said effective weed management was vital as weeds were the primary and most economically impactful pest in crop production.
He said crop yield losses to weed infestation were estimated to be about 9% globally, while in developing countries weed control could take up to 50% of farmers’ time.
In addition, he said invasive species of weeds were detrimental to native plants and could also be harmful to livestock and humans.
“Therefore, weed management is a critically important activity on agricultural and non-agricultural lands.
“Furthermore, current trends suggest that weed problems will worsen in the next 10 to 20 years due to global environmental change.“Future weed management efforts must be more innovative and holistic to gear towards sustainable intensification of crop production and environmental conservation,” he said.
He said there was a need to develop new tools and techniques for effective weed management in view of the manpower shortage in agriculture.
“This is especially important in large and commercial-scale farms. Technologies are already available on the application of robotics and machine vision for weed detection and removal.
“These digital technologies could revolutionise weed management leading to the year 2050,” he said when opening the 27th Asian-Pacific Weed Science Society (APWSS) conference here.
Giving an example, Uggah said hyperspectral imaging methods could now be used to detect weeds, regardless of the plant’s visibility or distinct leaf shape.
“Studies show that it can reliably distinguish closely-related plants, for example tomato and black nightshade,” he said.
Another promising technique, Uggah said, was the systems approach to develop smart machines for automated weed control.
In this technique, crop planting locations are mapped and a weeding robot can then access the map to distinguish the crops from weeds based on location.
Uggah said effective weed management was vital as weeds were the primary and most economically impactful pest in crop production.
He said crop yield losses to weed infestation were estimated to be about 9% globally, while in developing countries weed control could take up to 50% of farmers’ time.
In addition, he said invasive species of weeds were detrimental to native plants and could also be harmful to livestock and humans.
“Therefore, weed management is a critically important activity on agricultural and non-agricultural lands.
“Furthermore, current trends suggest that weed problems will worsen in the next 10 to 20 years due to global environmental change.“Future weed management efforts must be more innovative and holistic to gear towards sustainable intensification of crop production and environmental conservation,” he said.
Already a subscriber? Log in
Get 20% OFF The Star Digital Access
Cancel anytime. Ad-free. Unlimited access with perks.
