Commercial & research landscape for smart irrigation systems. A survey of commercial product offerings, research prototypes and approaches to smart irrigation. I also cover the why there is such a dire need to conserve water and increase yield.
Commercial & research landscape for smart irrigation systems
1. Commercial & Research Landscape for
Smart Irrigation Systems
Smart Irrigation Systems: Conserve water and enhance crop yields. “More Crop per Drop”
Muhammad Yaseen
(muhammad.yaseen@dfki.de)
Intern, Knowledge ManagementGroup
German Research Center for Artificial Intelligence (DFKI)
Kaiserslautern,Germany
Research Student, Koshish Foundation Research Lab
NED University of Engineering andTechnology
Karachi, Pakistan
2. Outline
▪ The Need forWater Conservation andYield enhancement
▪ Governmental and Global initiatives
▪ What Knowledge Management can Offer
▪ Commercial Products
▪ ResearchTrends
3. The Need forWater Conservation andYield enhancement
▪ Just about 2.5% of the Earth’s total water is freshwater that can be used for various
applications including irrigation. (US Geological Survey)
▪ On an average only 45% of the supplied water is used by crop (FAO).
▪ 15% is lost during conveyance
▪ 15% percent is lost in supply channels within the farms
▪ 25% percent is lost due to inefficient water management practices
▪ Population: 9.15 billionby 2050. Food production must increase by 70% (FAO)
4. Governmental and Global Initiatives
▪ European Initiative for Sustainable Development in Agriculture (EISA)
▪ Agriculture and Food Research Initiative (AFRI), USA
▪ WaterSense, USA
▪ Food and Agriculture Organization (FAO)
5. Common Goals: “More Crop per Drop”
▪ Increase Land andWater Productivity. “More crop per drop”
▪ Improve Crop yield
▪ Food security for growing population
▪ Reduce environmental impact of agriculture (Climate-SmartAgriculture)
▪ Provide farmers with actionable information about their crop and field
▪ Efficiently distribute and utilize irrigation water
6. What Knowledge Management can offer?
▪ Eliminate guesswork and hunches
▪ Combine raw data from different sources (sensor data, weather data, soil
characteristics)
▪ Provides growers with reliable and actionable information
▪ Efficiently use resources. (Water, Pesticides, Farm Equipment, Energy…)
▪ Visualizations and Data driven insights
7. A general framework for Knowledge Management
Machine
Learning
Neural
Network
Knowledge
Management
Field Sensor
Data
Expert’s
Knowledge
Crop and Soil
Requirements
Optimal Irrigation
Schedule
PrecisionAgriculture
Disease (Anomaly)
DetectionWeather Data
Facts based
predictions
8. Commercial Products
▪ 17companies and their product offerings were
examined.
▪ Different Market Categories
▪ Crop Health Management
▪ Farm Management and Agriculture Logistics
▪ Homeowners and small scale lawns / Commercial
landscapes
▪ Ornamental Plant production
▪ Shortlisted: 8companies
(considering only Crop Mgmt. and Farm Mgmt.)
10. Software
1. Data visualization (Graphs, Plots,
Summaries…)
2. Satellite and GIS info.
3. Predictive analytics
4. Remote control
5. Irrigation scheduling
6. Data driven insights
7. Real-time Notifications
8. …even APIs
Commercial Products:What do they
offer?
11. Comparison Matrix Of Commercial Products
Hardware on
Field
Software
Solution
Data Analysis
and KM
Weather or
Satellite Data
Rubicon Farm Connect Yes Yes Limited No
Crop Metrics Yes Yes Not Sure No
AgCoTech. Yes Yes Limited No
Agribotix Yes Yes Yes Yes
CropX Yes Yes Yes No
AgSmarts Yes Yes Yes No
FarmLogs Not Sure Yes Yes Yes
AquaSpy Not Sure Yes Yes No
Limited: Available features very limited as compared to other products in the matrix.
These companies offer solutions for crop health management, yield improvement,
farm automation and smart irrigation.Although the approaches are slightly
different.These companies are directly in competition with each other.
12. Research Outlook
▪ WSN for monitoring temperature, humidity, soil moisture, light intensity, greenhouse
gas levels.
▪ Arduino, GSM Module and MATLAB Fuzzy logic toolbox.
▪ Result: Irrigation decisions based on monitored parameters.
▪ Use of DecisionTree for crop water need prediction
▪ Parameters: Max/Min temperature, wind speed, humidity, rainfall, solar radiation, soil
type, crop type.
▪ Result: Authors report “74% accuracy”.The system was able to predict the water
usage and hence can help farmer make informed decisions.
Case # 1
Case # 2
13. Research Outlook
▪ Soil classification using 3 classification techniques. pH, ElectricalConductivity, Organic
carbon, amounts of P, Fe, Zn, Mn, Cu are learning parameters
▪ Algorithms used: Naïve Bayes, C4.5 (Decision tree), JRip . Rules for soil classification
were collected from soil testing lab.
▪ Result: Authors report a “91.9% accuracy” for C4.5 algorithm
▪ A two-year study ofTomato crops in Saudi Arabia.
▪ A test crop was grown with Intelligent Irrigation System (IIS). Parameters used: Solar
radiation, wind speed, humidity, rain, air temperature and ET.
▪ Result: Authors report “26% water savings”.
Case # 3
Case # 4
14. Research Outlook
▪ Sistema Irriga – Irrigation scheduling for 185,000 hectares of land in Brazil
▪ Parameters: Air pressure, temperature, humidity, wind speed, wind direction, rainfall,
solar radiation.
▪ Result: Daily irrigation recommendations. Forecasts for next 1-2 days.
Take home point!
▪ Irrigation can be scheduled to achieve efficient water usage by monitoring a set of
crop or plant parameters (Humidity,Temperature, Soil Moisture etc.)
▪ Case 1,2 and 3 acquired labelled data from expert’s knowledge and prior experience.
Case 4 and 5 have not mentioned how they acquired labelled data.
Case # 5
15. Relevant Conferences, Journals and Books
▪ InfoAg Conference
▪ European Conference on PrecisionAgriculture (ECPA)
▪ InternationalConference on PrecisionAgriculture (ICPA)
▪ InternationalConference on Computer and ComputingTechnologies in Agriculture
(ICCTA) and associated Journal of same name.
▪ PrecisionAgriculture Journal
▪ Data Mining InAgriculture
16. References
▪ TheWorld’s Water : http://water.usgs.gov/edu/earthwherewater.html
▪ How to feed the World in 2050:
http://www.fao.org/fileadmin/templates/wsfs/docs/expert_paper/How_to_Feed_the_World_in_2050.pdf
▪ http://crpit.com/confpapers/CRPITV121Khan.pdf
▪ https://www.researchgate.net/profile/Petraq_Papajorgji2/publication/225309627_A_survey_of_data_mining
_techniques_applied_to_agriculture/links/53ed09200cf2981ada11bb9c.pdf
▪ https://arxiv.org/ftp/arxiv/papers/1206/1206.1557.pdf
▪ http://www.ijetae.com/files/Volume5Issue4/IJETAE_0415_66.pdf
▪ http://www.cropj.com/marazky_7_3_2013_305_313.pdf
▪ https://books.google.de/books?id=DybQsneDddYC&lpg=PA7&dq=european%20conference%20on%20preci
sion%20agriculture%20proceedings&pg=PA19#v=onepage&q&f=true