Understanding and addressing the challenges
Reasons for low technology adoption are rooted across the layers of the agricultural technology stack.
- Wireless (mainly cellular) data connectivity is limited at the far edge spread across the vast area – farmlands are remote, mostly in rural areas. High bandwidth wireless endpoint connections are rare, due to a lack of infrastructure and investments. Deploying sophisticated edge gateways is cost-prohibitive as well, considering the range per gateway and area that needs to be covered.
- Minimal power budget may constrain sensor operating range, frequency, compute power, and life. Executing an edge workload – whether for sensing, rule-based control or compute, needs energy. Edge devices have typically used battery power. However, one needs to balance a large number of tiny sensors to be deployed in a vast area in scope with the battery specification to be used (considering the cost-effectiveness). While battery-less IoT devices have been recently developed with energy harvesting technologies like li-fi, it is yet to gain widespread adoption on scale and cost considerations.
- Challenging weather conditions like rain, wind, fog, heat, and snow impact the reliability and accuracy of the data stream in capture as well as transfer motions. Sensors capturing the outdoor environment parameters need to have a broad operating range w.r.t the ambient temperature, precipitation, humidity, visibility, and wind. This can potentially impact sensing as well as transmission accuracy.
As digital technology matures and becomes more accessible to corporations, co-operatives, and farmers, there is a push toward smart greenhouses, precision farming, and connected warehouses in the agricultural sector. As population and urbanization grow, we see a significant increase in food demand (70% growth by 2050) and reducing the acreage of agricultural land. The rising yield and prices of agricultural produce generate returns on technology initiatives for the majority of the farming entities.
Technology solutions to industry challenges
Technologies that are typically at the core of smart agriculture solutions include:
- IoT for edge connectivity and sensor telemetry data transfer over multiple short-range (open as well as proprietary) protocols.
- AI/ML for vision-based analysis of camera feeds from drones.
- Mobility for end-user, technician apps for data reporting, workflow automation.
- Cloud for sensor data management and analytics application platforms.
These technologies are leveraged in use cases like:
- Soil monitoring – Monitoring soil temperature and humidity periodically to keep them within the recommended threshold is essential for precision farming practices. You can achieve it using buried sensor probes that communicate the telemetry data to routers or gateways deployed on the farm. Managing watering and planting activities based on the present conditions and past trends helps ensure sufficient water for plant growth, effective delivery of water-soluble nutrients in manures and fertilizers, thus maximizing the yield. Soil temperature greatly impacts the mineralization of essential elements like nitrogen that is essential for root growth and respiration.
- Environment monitoring – Regular sensor-based monitoring of non-contact surface temperature around sprouted seeds is critical to ensure the right conditions for sapling growth. Even for the fully grown trees bearing flowers and fruits, ensuring certain temperatures around the trunk for healthy growth of the buds into flowers and subsequently to fruits becomes critical for maximizing the yield. Many crops and plants require ample sunlight for growth. You can deploy light sensors for assessing plots within the farm where such seeds are planted.
- Pest and weed detection – Vision technologies on autonomous vehicles like drones are used to detect the percentage of the weed coverage on the farm, identify beyond threshold plots, and accordingly manage farming manpower to remove the weeds. It is important to ensure that the input factors like water, essential nutrients in manures, fertilizers, and sunlight, are utilized for the crops. You can detect pests in the form of fungus and insects that reduce the crop yield, through deep computer vision technology practices like semantic segmentation, image classification, and object monitoring.
- Livestock monitoring – Sensor-based location tracking and geofencing along with the body vitals (e.g. temperature) monitoring ensure the accurate count and health of the livestock under management. Measuring and monitoring ambient conditions in the living areas help enable hygiene, safety, and overall wellbeing. Subsequently, dairy produce (milk, eggs), and meat from healthy animals ensure high agricultural throughput.
- Stored produce monitoring – Harvested produce often needs to be stored near the farm due to the weather, logistics, and market factors. The shelf life of the stored produce is determined by factors like external and internal temperatures of the storage bins and containers, humidity levels and airflow rates, and storage and light intensity. Monitoring these parameters and maintaining them at the levels specific to the stored harvested crop can be achieved using IoT, mobility, analytics, and cloud technologies.
The above use cases showcase that these digital technologies have the potential to transform the agricultural value chain across the crop lifecycle. If used at scale responsibly, these technologies can contribute to achieving multiple Sustainable Development Goals (SDGs) across the globe.
eInfochips, an Arrow company, is a leading product engineering services company with expertise across silicon to device to cloud, on multiple hardware and software technology platforms. We have experience in building connected solutions across industry verticals including agriculture, water management, and industrial manufacturing covering highly distributed field devices, sensors, and assets. To know more about our digital transformation services, talk to our experts today.
ABOUT THE AUTHOR
Nirupam Kulkarni works as Product Manager focused on eInfochips' IoT and AI product and solution portfolio. He has 11 years of technology experience across IT and analytics consulting, market research, product management and digital partnerships.