What can drones do for you: Precision Agriculture

INTRODUCTION

Ever since Unmanned Aerial Vehicles (UAVs), better known as drones, became commercially available in the early 1980s, their ever-increasing capabilities have led to new applications across a variety of industries. In just the past couple of years drone technology has advanced past the point of novelty. Drones are now a wise investment option, delivering valuable payback on initial cost.  Aided by recent evolutions in both drone technology and FAA regulations, one area that is quickly becoming revolutionized by drone technology is agriculture. As a result, drones have now been mainstreamed for farming, and these developments could not be timelier. With global population projected to reach over 9 billion by 2050, agricultural consumption is expected to increase by a massive 70%, a figure complicated by unpredictable weather patterns and natural disasters.

Drones can be a large part of the critical solution to this exponential increase in demand, along with closer collaboration between governments, tech leaders, and industry. Drones can assist famers in a range of tasks from analysis and planning, to the actual planting of crops, and the subsequent monitoring of fields to ascertain health and growth. As farms become larger and more efficient to meet this escalating demand, drones will prove invaluable in precisely managing a farm’s vital operations. 

In an attempt to demonstrate the value of drones along with related artificial intelligence in precision agriculture and in collaboration with MINADER, and IRAD, we performed a small campaign at a seeding farm belonging the Menoua division in the west province of Cameroon. Multiple flights and analysis were performed.

The following results are from survey of a 1.4 hectares corn field with a actual flight time of 5 minutes and detailed analysis of about 24 hours.

The density of the plants determines the plant population in any given area. The closer the spacing, the more plants there will be per unit area. The recommended number of maize plants per hectare varies from 36,000 to 60,000, depending on the environmental yield potential and hybrid. High plant populations are appropriate for early-planted crops under high rainfall or irrigated conditions where management is of a good standard. Lower plant populations should be used under dry land conditions, especially in drought prone areas. Maize faces several constraints during cultivation and fast/early identification of these constraints are very important for timely implementation of measure to ensure maximum yield. This can be done using drones which are fast and effective, providing accurate and timely information.

 

Plant Count Analysis

The area under study is a 1.4 hectare maize farm at a growth stage of V3. Based on count data from our AI analysis, the total number of maize plants counted was 20,941 as indicated in figure. This number is low compared with the recommended number of maize plants for a hectare. This is due to several factors such as: death of plants due to stress and presence of death corn on almost an half of the area. The healthy corn area is exactly around 0.54 hectare.


I am satisfied with my meeting with Skyvue Solutions today. I am impressed to see how these new technologies make it possible to accurately determine the area of cultivated land and accurately obtain the number of plants

Mr Dongou
MINADER Representant

Plant Stress Analysis

Further analysis was carried out for general farm and plant characteristics to account for the low number of maize plants in the study area. Soil characteristics such as nitrogen content, water content, insect pests can account for major stress faced by crops. Subsequent figures gives a clear picture of possible reasons.

General stress level of crops was estimated at 97.14% with only 2.86% of plants growing without any stress 

Nitrogen Analysis

Further analysis showed that 22% of the farm had low levels of nitrogen which might have contributed to the stress of crops, leading to poor vegetative growth, chlorosis, stunting and consequently poor yields.

Water Logging Analysis

Potential water logging for the area as depicted was 9 % of the entire field as shown in the figure. While 90.71% of this field, about 1.27 Hectare is not gorged. A plausible explanation to this could be the high permeability of the soil which results from its high sand content. Water logging may exert stress for the plant by promoting anaerobic conditions thus affecting some physiological processes requiring oxygen such as respiration. Some nutrients may become unavailable to the plant under water logged conditions. Roots may also rot under such conditions. Overall, the plant growth will be affected negatively. The low total number of plants in the field can be attributed to water logging. 

Pest Analysis

As shown in this figure, the farm is healthy. 

Weed analysis

As showed in the figure, further analysis showed that the plot is overgrown with weeds at close to 70%. This has a direct impact on plant stress and yield.


I am impressed by this tool that will revolutionize our agriculture. I trust you to make this available to many farmers in Cameroon

Mr Temgoua
MINADER Representant

Recommendations

Based on the results, the percentage of water logged areas compared to field size was not significant and therefore could not be a major stress factor in this field. However, with the low Nitrogen content, we recommend fertilization (top dressing) with Nitrogen based straight fertilizers (e.g ammonium nitrate, urea) to improve on the vegetative growth of crops; which is important for reproduction (formation of cob/ear). Given the signifiant presence of weeds with little to no insect infestation, we also recommend the targeted use of herbicides over the most affected zones. 

Conclusions

Traditional crop counting methods are very tedious, time consuming and generally quite innacurate.
Using traditional methods, it’s also impossible to accurately evaluate the Nitrogen level of the plants as farmers are often limited to making assessments based on the leaves going from green to Yellow at which point it’s usually too late to remediate.
In similar fashion, traditional methods of checking for insect infestation rely on visual Detection of insects or their lavae on plants. Again this is usually very hard to see on one plant, let alone an entire farm.

Drones however, can perform all those tasks quicker, safer, and more accurately. For instance, in less than 30 minutes, our drone flew over the entire farm and processed all the data in less than 24 hours to provide us with all the analysis cited above. Namely: plant count, weed/plant stress analysis, insect infestation,… . In addition to the effectiveness of the drone, it also gives us a more focused analysis which allows us to be more precise in our application of remedies. In this case, rather than spraying the entire farm upon detection of weeds on one side of it, we are able to focus our effort only on the areas called out in the map. By so doing, we save time, and money which ultimately hits the bottom line of the farmer.

Contributors: Gabriel Kamdem – Agropedologist and researcher at IRAD; Ethelyn Forchibe – Agropedologist

https://youtu.be/Vdn00-03IfA
Rate this post
5/5

Leave a comment