Integration of a wireless sensor network and a participatory soil monitoring system for smallholder agriculture PhD proposal qualifier presentation Amsale Zelalem Supervisors: Dr. Javier Morales, Dr.ir. Rolf de By, Dr. Yaregal Assabie February 26, 2019
Outline Motivation Methodology: High level sketch Case study Conclusion Execution plan February 25, 2019 2 / 24
Overview Motivation Methodology: High level sketch Case study Conclusion Execution plan February 25, 2019 3 / 24
High-end agriculture February 25, 2019 4 / 24
High-end agriculture ◮ Digital information and ◮ technology support ◮ Adequate production: ◮ quality and quantity ◮ Nature resilience ◮ ◮ Mechanized and large ◮ farms ◮ Regular follow-up and ◮ advises February 25, 2019 4 / 24
Low-end agriculture February 25, 2019 5 / 24
Low-end agriculture ◮ Smallholders and ◮ subsistence-oriented ◮ Intuitive farming ◮ ◮ Very low technological ◮ base and advisory support ◮ Deficit production ◮ February 25, 2019 5 / 24
Key drivers: Concept diagram February 25, 2019 6 / 24
Gaps ◮ Extreme scarcity of up-to-date soil data at farm field level ◮ ◮ Inefficient and intuitive use of inputs ◮ ◮ Soil data collection is resource-intensive and difficult ◮ ◮ Advisory systems lack crop-specific soil requirements ◮ ◮ Deficiencies remain undetected until too late ◮ February 25, 2019 7 / 24
Intervention ◮ An integrated solution of natural, bio-physical, cultural and ◮ technological flavors are needed February 25, 2019 8 / 24
Intervention ◮ An integrated solution of natural, bio-physical, cultural and ◮ technological flavors are needed ◮ Implement a (near) real-time, robust, usable, rapidly ◮ deployable and affordable soil data collection and analysis tool at farm field level ◮ Create a platform for improved information flow from ◮ multiple sources to improve agriculture yield predictions and projections February 25, 2019 8 / 24
Soil ◮ A source that supplies plants with water and nutrients ◮ ◮ Every crop needs a special composition of the soil to ◮ achieve its potential ◮ Moisture links drought-climate-vegetation ◮ ◮ Nutrients determine yield quality and quantity ◮ February 25, 2019 9 / 24
Overview Motivation Methodology: High level sketch Case study Conclusion Execution plan February 25, 2019 10 / 24
From ground to ground Data Sources Network Server Knowledge base Sensors continually monitor the soil Application Server February 25, 2019 11 / 24
Methodology: system decomposition Integration of WSN and participatory soil data acquisition platoform Usability & impact assessment Soil moisture aquisition Site selection Network design Farm identification Input use Device calibration & configuration Crop identification Network deployment Empirical analyis Evaluation Data acquisition Data validation Crop physiology Agriculture expert system Particpatory soil analysis Recruit volunteers Set science-kit Hub Pre-processing Integration Inference design Data acquisition System deployment Knowledge base Evaluation Evaluation Knowldege produce February 25, 2019 12 / 24
Usability and impact assessment: Hypthesis-driven approach Process Input Investigation Define area of interest Crop data Yield data Empirical & Soil data qualitative Define temporal Triangulation Input use evaluation resolution Imagery Sen2Agri Satellite Atmospheric Knowledge produce correction Repositories Crop type Land mask Experts Biophysical VIs indicators February 25, 2019 13 / 24
Soil moisture aquisition: Wireless Sensor Network Backhaul Layer Network server TCP/IP MQTT/ Publish Brok roker/to /topic TT TTN Cellular TCP/IP- Gateway y WiFi MQTT/ subscribe LoRaWAN CN CN 1 Application Layer Application CN n CN interface ECH2O 5TM ECH2O 5TM Persistent Store Sensing Layer Application server EN EN 2 2 EN EN 1 1 EN EN 3 3 EN EN n n ECH2O 5TM ECH2O ECH2O ECH2O 5TM ECH2O 5TM February 25, 2019 14 / 24
Participatory soil analysis: Digital citizen science System interface User-end Back-end Observation & Survey Pre- design analysis processor Science Hub System design & Kit selection & Repository setup guideline design Implementation Volunteers’ Artifact recruitment & selection training February 25, 2019 15 / 24
Agriculture expert system: Artificial Neural Network Expert System Knowledge base Wisdom/ decision support Existing data - Make - Define rules Yield Land inference - Set relevance estimation * - Conceptualize evaluation Infer - Learn patterns * respond - Test inference - Compute error - Back propagation - Commit - Adjust relevance * Organized Input/Data * pre-processed Learn Information Internal External External - WSN-IoT - Experts Land factors Social factor External External - Participatory soil - Literature analysis - Satellite - Farm practices - Weather stations - Physical - Input use - Chemical * Facts * Environmental External Crop factors External factors - Soil - Location - Crop type -Crop - Elevation - Crop requirement - Climate - Precipitation - Yield patterns - Farm mgt. - Temperature - Weather stations - Humidity - Topography February 25, 2019 16 / 24
Overview Motivation Methodology: High level sketch Case study Conclusion Execution plan February 25, 2019 17 / 24
Study area: Beshilo basin ± ± B Gidan A Meket Lay Gayint Guba Lafto Esite Wadla Tach Gayint Dawunt Delanta Amba Sel Simada ◮ Two districts: Kutaber and Dessie ◮ Kutaber Mekdela Tenta Sayint Dessie Zuriya Dessie Zuria Legambo Bounary Were Ilu 1:23,173,505 Rivers 1:1,893,707 ± ± C D ◮ Total area of 2206 km 2 ◮ ◮ More than 300,000 population ◮ Dega 1,202 - 1,700 2,700 - 3,200 Kolla 1,700 - 2,200 3,200 - 3,700 Weina Dega 2,200 - 2,700 >3700 Wurch 1:2,052,626 1263744 .095916 517912 .118082 527912 .118082 537912 .118082 547912 .118082 557912 .118082 567912 .118082 577912 .118082 587912 .118082 597912 .118082 1263744 .095916 ◮ Rurally dominated and E ◮ ± 1253744 .095916 1253744 .095916 agriculture-dependent livelihood 1243744 .095916 Kuta Ber 1243744 .095916 ( ! Kutaber Alanshana Werkaya ◮ Severely damaged soil and high Liwicho 1233744 .095916 1233744 .095916 ◮ Goro Mender Dese prevalence of food insecurity ( ! 1223744 .095916 1223744 .095916 Kolamote 020 Dessie Zuria Degamote 021 1203744 .095916 1213744 .095916 1213744 .095916 Source:EthioGIS2 Kebeles Atnt Mesberiya 030 Sites 1:555,769 ! ( Town 517912 .118082 527912 .118082 537912 .118082 547912 .118082 557912 .118082 567912 .118082 577912 .118082 587912 .118082 597912 .118082 February 25, 2019 18 / 24
Collaborations February 25, 2019 19 / 24
Collaborations ◮ Research collaborations ◮ Application of digital image processing for parcel delineation using remote sensing and GIS Design and implementation of standards for heterogeneous data acquisition, integrations and retrieval February 25, 2019 19 / 24
Collaborations ◮ Research collaborations ◮ Application of digital image processing for parcel delineation using remote sensing and GIS Design and implementation of standards for heterogeneous data acquisition, integrations and retrieval ◮ Logistics collaborations ◮ Equipment are offered by Dr. Rogier from WRS Network back-end configuration & maintenance assist from Mr. Joreon:LISA (ITO) Grant proposal to USAID: ” Building civic participation, good governance, and resilient communities” February 25, 2019 19 / 24
Collaborations ◮ Research collaborations ◮ Application of digital image processing for parcel delineation using remote sensing and GIS Design and implementation of standards for heterogeneous data acquisition, integrations and retrieval ◮ Logistics collaborations ◮ Equipment are offered by Dr. Rogier from WRS Network back-end configuration & maintenance assist from Mr. Joreon:LISA (ITO) Grant proposal to USAID: ” Building civic participation, good governance, and resilient communities” ◮ Possible coherence with parallel PhD works of the ◮ EENSAT project February 25, 2019 19 / 24
Overview Motivation Methodology: High level sketch Case study Conclusion Execution plan February 25, 2019 20 / 24
Work significance and contribution February 25, 2019 21 / 24
Work significance and contribution ◮ Real-time, affordable, and participatory field-level soil data ◮ acquisition system ◮ Integration of multiple data sources to complement ◮ agricultural decisions ◮ Produce baseline for satellite data calibrations & ◮ validations ◮ Fill farm-level information gaps to precision farmings ◮ February 25, 2019 21 / 24
Work significance and contribution ◮ Real-time, affordable, and participatory field-level soil data ◮ acquisition system ◮ Integration of multiple data sources to complement ◮ agricultural decisions ◮ Produce baseline for satellite data calibrations & ◮ validations ◮ Fill farm-level information gaps to precision farmings ◮ February 25, 2019 21 / 24
Overview Motivation Methodology: High level sketch Case study Conclusion Execution plan February 25, 2019 22 / 24
Recommend
More recommend