The SCRI-MINDS Project

Wireless Sensor Networks for Decision Irrigation Management

George Kantor 1 and John Lea-Cox 2

1 Robotics Institute

Carnegie Mellon University

2 Dept. of Plant Science and Landscape Architecture University of Maryland

JLC@umd.edu

Funding provided by USDA-NIFA-SCRI Award no. 2009-51181-05768

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

1. What are the Issues?

2. Hardware and Software Development

3. Grower Partner Implementation, Results

4. Economic Impact, ROI

5. Challenges

Presentation Outline:

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Because very few growers are monitoring practices

Why Sensor Networks?

Water management is the key to nutrient management and optimizing growth

Growers typically won’t change practice unless you convince them that it will improve productivity or profitability

We need to move from precision irrigation to precision + decision irrigation management

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Fertilizer

Surface runoff

Substrate

components

Water and Nutrient Uptake

Nutrient

leaching

pathogens / water

Irrigation Water

Application

Plant Yield

Water Management is the Key to Many Issues

Courtesy of Dr. Jim Owen, Virginia Tech University

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

University of Maryland:

• John Lea-Cox (PM, Micro-scale

Modeling – Nursery, Extension Outreach)

• Andrew Ristvey, Steven Cohan (Green Roof)

• Erik Lichtenberg (Economics – Private Benefits)

Carnegie Mellon:

• George Kantor, David Kohanbash

(Next Generation Core Software Development)

Cornell:

• Taryn Bauerle (Microscale – Root Environments)

UM-Center Environmental Science:

• Dennis King (Socio Economics – Public Benefits)

Decagon Devices: • Todd Martin, Colin Campbell Lauren Bissey

(Next Generation Hardware Development, Core Software)

Antir Software:

• Richard Bauer

(Crop Modeling Software,

Core Software )

Colorado State:

• Bill Bauerle, Mike Lefsky, Stephanie Kampf

(LIDAR, Hydrology, Macroscale modeling – Nursery)

University of Georgia:

• Marc van Iersel, Paul Thomas, John Ruter, Matthew Chappell (Microscale modeling – Greenhouse, Extension and Outreach)

SCRI-MINDS: Teams and Working Groups Sensor Networks

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Efficient; Information must be easy to interpret

(5-minute decision window)

Systems must be robust (low maintenance)

Technology needs to be cost-effective (short ROI)

Successful Implementation:

Sensors need to be accurate (plug ‘n play)

nR5-DC Node

nR5 Node – This wireless node allows us to both monitor sensors and control irrigation events, based on sensor readings

Node measures data every minute and then logs the data at an interval specified by the user (1, 2, 5, 15, 30, 60 minutes etc.)

Monitoring Mode: Batteries logging at 15 minutes typically last 12+

months

Control mode: Batteries are lasting 4-6 months, depending on the # irrigations initiated per day.

New nR5-Control Node: (Field tested during 2012)

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Sensorweb Software: Enables remote and/or automatic control of irrigation schedules, via a customized web-based interface

Sensorweb: Macro-Scheduling Tool

Maple Block

Dogwood Block

Allows for real-time monitoring and adjustment of irrigation events, for blocks of times during the day, using sensor-based or schedule-based control

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Allows a “time-out” for sensors to measure between pulse events, reducing leaching fractions (and nutrient loss) to minimal amounts

Sensorweb: Micro-pulse Tool

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Augments Scheduling and Micro-pulse tools with soil moisture sensor feedback: irrigation is disabled with set-point is exceeded

Sensorweb: Local Set-Point Irrigation

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

12V DC latching solenoid + 1” valve

Bleed valve

Flow meter

nR5-DC Node

nR5-DC Node – Integrated with a flow meter and controls a 12V solenoid valve

Allows us to control and measure water applications in remote fields where there is NO electrical power

12V Solenoid and Flow meter Integration

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Current BMP: Timed Cyclic Irrigation

4 x 6-minute Irrigation Events = 164 Gals / day

Monitoring Block: 3 to 4 timed cyclic irrigation events per day

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Future Management Practice?

2-minute Irrigation Events

1-2 Irrigations per day @ 21 gal / 2 min pulse

Control Block: Reduced (2-min) pulse events PLUS a reduction in number of total irrigation events (based on substrate water content)

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Dogwood Monitoring vs. Control

Water Use: April – October, 2012

Irrigation

Method

Total Water Use

(Gals / Row)

Average Water

Application

(Gals/ Tree /Day)

Av. Efficiency

(Timed vs.

Control)

Water Savings

(Control vs.

Timed)

Grower: Timed,

Cyclic 28,334 0.922

0.371 2.69 Sensor: Setpoint

Control 10,521 0.342

Saving Water in Three Ways: Supply and Demand

1. Micropulse: Smarter delivery, through reduced irrigation duration

Outcome: Oftentimes, only a single 2-minute pulse is required

2. Environment: Sensing plant water demand, as a function of

changes in environmental conditions

Outcome: Reduction in total number of irrigations per day, based

on evapotranspiration demand, (light, temp/RH and rainfall)

3. Growth: Sensing plant water demand, as a function of plant

growth Outcome: Irrigation scheduling is based on plant growth (increased or decreased demand), optimizing root water status

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Implementation, Return on Investment

McCorkles Nursery (GA)

14-month production cycle collapsed to 8-month

30% loss to Disease reduced

to virtually zero

> 100 Million gal reduction in total water use / year

Economic Gain = $1.06 / ft2

(total net revenue = $20,700 for crop)

ROI < 3 months for $6,000 network

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Implementation, Return on Investment

Hale and Hines Nursery (TN)

Sensor-controlled Irrigation reduced water use by 63% = $$$ in labor (TBD)

Growth of trees was equivalent

> 43 Million gal reduction in total water use / year; $6000 in pumping costs

Translated to CA, net savings in water cost (@ $750/ acre ft) = $100K per year

ROI < 3 months for $25K network

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Challenges

Scaling networks to large acreages

Nurseries are site-specific in their needs (problems)

Many species with specific needs (will probably require a “indicator species” approach)

Support Issues (Consultant network)

for specific regions (environments)

and specific crops

Apply to other specialty crops: need

partnerships with domain specialists

and growers.

It’s a big country!

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Smart-Farms.net — Managing Irrigation and Nutrition via Distributed Sensing

Project Information at http://smart-farms.net

Funding provided by USDA-NIFA-SCRI Award no. 2009-51181-05768