View
226
Download
4
Category
Preview:
Citation preview
Multisensory Agricultural Monitoring Platform
PI: Ketan Rajawat (EE)
Co-PI: Rajesh Hegde (EE)
Multi modal PBX/Voice server
Database Server (PC Debian GNU/Linux)
SMS Gateway
Get Crop Status via SMS and Voice Call
Get required disease infomation and Store Call information
Make a call and follow instructions
Getting crop infestation knowledge data from
central disease server or distributed servers
Web Server (PC Debian GNU/Linux) Distributed
Servers
Goals
• Agricultural practice in India is characterized by a large pool of local knowledge
• Communities thrive on local wisdom about dealing with local environment, infestation, etc.
• However: central planners often miss this data
• Bridge the gap between: technical knowledge & longitudinal farm experience
• Goal: agricultural monitoring, data collection, presentation, trend mapping, visualization
Objective
• Multi-sensor agricultural monitoring and information dissemination platform
ZigBee/WiFi gateway
ZigBee/GPRS
gateway
Sensor nodes
Cell phone
network coverage
“Data Mule”
(Farmer)
WiFi communication range
Data Mule’s
movement path
Deliverables
• Identification and deployment of affordableand usable sensors
• Architecture: mobile data collection agent
• Backend database + data analytics codebase
• Web portal with maps & alert services
Motivation
• Enable epidemiological studies at central level
• Allow data scientists to spot trends
• General early warning signals on movement of infestations and impact of weather conditions
• Overall: enhancing P2G and G2P
Novel Aspects
• Energy & bandwidth efficient data collection1
• Novel strategies for distributed storage2
• Succinct & meaningful data visualization3
• Filtering, data cleansing, & outlier rejection• Path planning• Beyond
– Predictive models for diseases– Real-time query resolution
1. K. Rajawat, A. Cano, and G. B. Giannakis, “Network-compressive coding for wireless sensors with correlated data,” IEEE Transactions on Wireless Communications, vol. 11, no. 12, pp. 4264-4274, Dec. 20122. Lakshmi J Mohan, Udaya Parampalliy, Ketan Rajawat, and Aaron Harwood, “An optimization design framework for repair efficient storage codes in geo-diverse clusters,” submitted to IEEE Globecom 20173. K. Rajawat and S. Kumar, "Stochastic Multidimensional Scaling," IEEE Transactions on Signal and Information Processing over Networks - Special Issue on Distributed Information Processing in Social Networks, vol. 360-375, no. 2, pp. June 2017.
WP1: Sensing technologies
• Investigation & testing of sensing technologies– Agricultural sensors for use in Indian context
– Field testing of low-cost sensors
Air temperature Leaf wetness
Air relative humidity Stomatal conductance
Solar radiation Soil temperature
Wind velocity Soil moisture content
Rainfall Soil pH
Soil salinity Soil bulk density
Soil hydrolic conductivity Cameras
WP2: Architectural issues
• Design of sensor-gateway data collection
– Energy efficient + low power protocols
– Sleep-wake protocols for energy conservation
• Lab testing of appropriate protocols
– Implementation on test bed in lab
• Reliability and stress testing
– Lifetime maximization
– Operation under harsh conditions
WP3: Data Mule
• Algorithm design for low-cost data collection
– Data compression
– Path planning
– Distributed storage
• Implementation of data fetching protocol
– At WSN lab in EE dept.
– Ensure robustness to delays and losses
• Field testing
S. Kumar, R. Jain, and K. Rajawat, "Asynchronous Optimization Over Heterogeneous Networks via Consensus ADMM," IEEE Transactions on Signal and Information Processing over Networks, vol. 3, no. 1, pp. 114-129, Mar. 2017.
WP4: Backend Design
• Remote database server
– Pulling data from data mule
– Pushing queries to sensors through data mule
• Data analytics
– Storage, filtering, interpolation, prediction, outlier rejection, data cleansing, denoising
• Visualization & Mapping
• Real-time algorithms for query processing
WP5: Web Portal
• Software specifications for web portal
• Multimodal information dissemination tools– Text & map from backend
– Automated audio text messages
• Design & testing
• User-aided testing
• WP6: Integration & field testing
Deliverables• Year 1: Investigation and field
testing of agricultural monitoring sensors
• Year 2: Implementation and testing of various routing, data collection, data analytics, and visualization algorithms
• Year 3: Field testing of the routing and data collection algorithms
• Year 4: Backend ready, allowing collection, storage, and processing of data
• Year 5: Web portal online and ready for deployment
SWP Year 1 Year 2 Year 3 Year 4 Year 5
1.1
1.2
2.1
2.2
2.3
3.1
3.2
3.3
4.1
4.2
4.3
4.4
5.1
5.2
5.3
5.4
6
Specific Technology outcome at Y1
• We expect to have completed the following at the end of the first year:– Comparative analysis of agricultural sensors
available in the market and their usability in the Indian context (field testing)
– Various algorithms for routing and data collection (SWP2.1 and SWP3.1)
Thank You
Recommended