/ THE FUTURE OF FARMING: SMART, AUTOMATED & SUSTAINABLE
By Valeri Oliver, Managing Editor, Avnet Content Strategy
AI at the edge, advanced sensors key enablers of drone technology for ag
The most thrilling design challenges might just be in our fields, orchards and vineyards.
By some accounts, agriculture sits on the verge of transformation not seen since the Industrial Revolution (1760-1840) when the cotton gin, reapers and threshers were introduced and made farming less labor intensive and more productive.
Drones equipped with on-board edge AI processors along with multispectral and hyperspectral imaging will enable autonomous assessment and action, which means farms will be more resilient to the unpredictable challenges of weather, pests and disease.
We're talking drones that don't just fly–they think, analyze, act autonomously and respond in swarms.
Why design engineers are on the ground floor of this innovation is no surprise.
Half of the world’s habitable land is used for agriculture. And, the global population has more than doubled since the mid-20th century. To meet the demands of a rapidly growing population on a planet with finite land resources, improved crop production is essential.
Design engineers are leading the way in harnessing and creating the next generation of autonomous systems to address the world’s growing food challenges, improve productivity and reduce the environmental impacts of farming. The future of farming is smart, automated and sustainable, all exciting design challenges.
We're talking drones that don't just fly–they think, analyze, act autonomously and respond in swarms.
Sensors that see the invisible
Sensors that can “see” beyond the human visible spectrum form the core of future agriculture drone technology.
Already well-established, multispectral imaging captures data in the near-infrared range to monitor plant health, water stress and disease presence.
What’s next? Hyperspectral imaging, or the ability to “read” hundreds of light bands across the electromagnetic spectrum to unlock insights. For example, drones equipped with these sensors could see subtle chemical plant changes before any visual stresses appear. Nitrogen deficiencies or fungal infections could be detected at the molecular level. That gives farmers the ability to act before problems are apparent.
But acting before problems escalate isn’t the only advantage. This type of sensor will reduce the need for broad chemical applications on fields since targeted fixes can be applied instead. That’s better for the environment, less costly for the farmer and results in better yields.
The next leap forward involves quantum sensors combined with predictive modeling. Using principles of quantum mechanics, quantum sensors could measure crops with precision down to molecular levels. Mounted on drones, these sensors when paired with predictive modeling could forecast droughts months early so irrigation patterns can be changed.
Quantum sensors could help predict drought months earlier than current technology.
MULTISPECTRAL/HYPERSPECTRAL COMPARISON
The multispectral approach involves images taken in several spectra. In hyperspectral imaging, the images are taken in may different spectra.
Source: Edmund Optics
LiDAR & advanced 3D mapping
These are also key ag drone technology enablers.
LiDAR (Light Detection and Ranging) uses laser pulses to capture details even in forests or on slopes where GPS struggles. LiDAR is set to make its mark in agriculture by offering ultra-precise topographical data as it has done already in forestry and mining.
The benefits for farmers include better irrigation management, more accurate planting strategies, and optimized land use for better yields.
MARKET OUTLOOK
The AI at the edge market is experiencing significant growth and adoption across various industries, including agriculture. According to Statista, the global intelligent industrial edge computing market is projected to reach $30.75 billion by 2025, with a compound annual growth rate (CAGR) of 18%. Edge AI processor shipments were expected to rise from 340 million units in 2019 to 1.5 billion units by 2023.
Statista also forecasts the global market size of smart agriculture, which includes drones, is expected to more than double from $15 billion in 2022 to $33 billion in 2027.
(NOTE: Statista is a German company that specializes in data gathering and visualization.)
AI at the edge: Where autonomous farming happens
Sensors are only as powerful as the algorithms that can interpret their data. The rise of AI and machine learning will be a game-changer, enabling agricultural robots and drones to make real-time, intelligent decisions without the need for constant human oversight.
The ability to collect all this new data requires a concerted effort on the processing side of things. AI also will play key roles in the future of farming with on-board processors handling data locally—sometimes known as inferencing at the edge or AI at the edge.
Instead of relaying the data back to the cloud for analysis, the processing will be handled locally at the drone level. Better yet, drones can act on that data, even delivering precision doses of fertilizer where needed.
DESIGN CHALLENGES: Because these drones are used in harsh agricultural conditions, design engineers must balance processing power with energy efficiency to minimize the impact of distorted, incomplete or “noisy” data. They should also be ready to develop algorithms that can make complex decisions despite operating in a harsh environment.
SWARM INTELLIGENCE
Robotics systems working together in the form of drone “swarms” will tackle modern ag’s scale and complexity.
Collaborative drones would cover large areas efficiently and distribute tasks dynamically. For example, one drone maps the terrain while another analyzes crop health. Co-working robots could then use that information and spray crops only where needed.
DESIGN CHALLENGES:
Reliable, secure communication protocols are critical to swarm systems. Engineers must also develop algorithms that enable drones to dynamically adjust behavior based on real-time conditions and shared data.
Using hyperspectral imaging to detect grapevine water stress
by Matthew Malcolm, California Ag Network
Groups of drones can work together autonomously to plant seeds, spray pesticides or pollinate flowers.
THE BOTTOM LINE
OEMs can position themselves at the forefront of these trends by acting now to harness the potential of advanced sensors, AI and autonomous systems in agriculture as they are keys to meeting the world’s food needs.
Market landscape
While private research firms provide specific market projections for agricultural technology, we can also gain insights from publicly available data:
GLOBAL AGRICULTURAL PRODUCTION: The global gross production value of agriculture was approximately $5.17 trillion in 2022. Source: Food and Agriculture Organization of the United Nations (FAO) LEARN MORE
U.S. FARM INCOME: Net farm income in the United States is forecasted to be $140 billion in 2024, down 4.4% from 2023. Source: U.S. Department of Agriculture, Economic Research Service LEARN MORE
GLOBAL AGRICULTURAL LAND USE: As of 2020, approximately 4.85 billion hectares (about 37% of the world's land area) were used for agriculture. Source: World Bank LEARN MORE
PRECISION AGRICULTURAL ADOPTION: In the U.S., farmers are increasingly using a variety of newer digital technologies. LEARN MORE
Examples of emerging precision agriculture technologies
Source: U.S. General Accounting Office summary of literature, interviews and agency documentation. January 2024 Technology Assessment on Precision Agriculture
ABOUT THE AUTHOR
Valeri Oliver
Managing Editor, Avnet Content Strategy
Valeri has more than 25 years of experience writing about a wide range of technology subjects for public companies, universities and publications. Before that, she was a newspaper reporter covering everything from city and state government to nuclear power.