Development and Applications of Unmanned Aerial Vehicles: Challenges and Opportunities

Yanbo Huang

Research Ag Engineer, Lead Scientist, USDA-ARS

Adjunct Professor, Texas A&M University, Mississippi State University, Delta State University

Unmanned Systems

• The operation of the systems is controlled either under a human operator through electronic control or autonomously by onboard computers

• Space-borne – Satellite control and operation

• Airborne – airplane and unmanned aerial vehicle (UAV) flight control

• Ground-based – Unmanned ground vehicle (UGA) control and operation

UAV-UAS-Drone

• Unmanned aerial vehicle (UAV) – Science/Engineering

• Unmanned aerial system (UAS) - Industry

• Drone – Media/Military

Any aircraft without a human pilot aboard

• Autonomous (Automated)

• Remotely piloted (Manual)

UAV History

• Began in the early 1900s

• Used in military in World War I (USA) and advanced in World War II (UK, USA)

• Civil use of UAVs follows up after the Cold War

UAV Market in the World

• USD 52.3 billion in 2025 (based on 2017)

• USD 67.3 billion in 2024 (based on 2014)

UAVs in USA

• 770,000 FAA drone registrations in March 2017

• Over 1 million in January 2018

UAVs in Europe

• USD 0.3 billion projected commercial drone revenue in 2025 (based on 2015)

UAVs in Japan

• Air robots

• First country to use and commercialize UAV technology for plant protection

In China

• Military• First successful unmanned flight in1959

• Established the first university UAV research department in 1962

• In 1964 and 1979 approved for research institute of UAV technology

• Institute of Remote Sensing Applications, Chinese Academy of Sciences in 2000s• Space missions

• Disaster monitoring

• Now – 10% world market share and 75 billion RMB in 2025 (based on 2018)

UAV Components

• Flight control (brain of UAV)

• GPS

• Data acquisition (IMU to maneuver)

• Autopilot (automated/autonomous)

• RC control (remotely piloted)

UAV Characteristic Parameters

• Weight (include battery or fuel)

• Payload

• Duration

• Ceiling

• Size/dimension

• Speed

• Operating temperature

• $$

UAV Variety

• Fixed wing

• Helicopter

• Multirotor, popularized in early 2010s

Wide variety of shapes, sizes, configurations, and characteristics.

Applications of UAVs

Commercial and consumer (4:6)

• Photogrammetry and remote sensing

• Precision agriculture• Low-altitude remote sensing

• Chemical application

• Forest and ecosystem

• Urbanization surveying and mapping

• Media coverage and film production

• Homeland security

• Hobbyist/DIY

• Others – your imagination and creativity

UAV Flight MissionMission

Planning

Flight

Mission

Complete?

Back to Origin

Stop

Mission?

Upload Data

to Cloud?

Upload and

Data

Processing

Download Data

from Memory

Card

No

Yes

No

Yes

Yes No

UAV Products

Consumer grade

Industrial grade

Sensors Used on UAVs

Multispectral

RGB Thermal Hyperspectral

Also portable LIDAR and SAR systems and the system for

Oblique Photography

Open Source Sensors on UAVs

Infrared temperature (IRT) sensors for temperature of

• Ambient air

• Soil

• Canopy

Infrared and time of flight (ToF) distance sensors

(Ultrasonic, LIDAR) for:

• Plant height

• Biomass

Optical sensors – Ultraviolet, RGB, Infrared

…

UAV Data Processing

• Orthomosaic images• Spectral images: RGB, multispectral, thermal,

hyperspectral

• Point clouds• Digital Surface Model (DSM)• Digital Terrain Model (DTM)• 3D modeling – Urban model, crop plant

height, biomass• Multitude of visualization, animation,

rendering and mass customization

UAV Data Products

• Can fly at very low altitude for target-specific operation with very high resolution image data (cm – mm/pixel)• Airborne remote sensing – 0.5-1m/pixel• Satellite

• High – 1-15m/pixel• Medium – 30-50m/pixel• Low – 250-1,000m/pixel

• Negligible atmospheric interference but cloud shadows still a problem

• Can flexibly fly into hilly, mountainous and valley areas

• 3D point cloud data processing from stereo vision with SfM

• Near-ground truth data measuring without restriction of field conditions

Advantages of UAV Remote Sensing

• Remote sensing

• Pesticide application

• Japan and China

• Australia and New Zealand

• South America

• United States

• Europe

• Others

UAV-Based Systems for Precision Agriculture

Zonal

Analysis

Pre-Remote

Sensing

Prescription

Map

Spray

ApplicationPost-Remote

Sensing

Satisfactory?

Meteorological

Observation

Application

Technology

Spray

Models

Engineering

ExperimentsComplete

Yes

No

• Dual Camera Systems

• Plant height estimation

• Crop yield estimation

• Crop stress

• Pest management

• Crop herbicide injury

• Herbicide-resistant weeds

• Nutrient management

• Irrigation management

• Crop breeding and phenotyping

• Crop harvest management

Research, Development and Agricultural Applications of UAV-Based Systems

y = 1.2x - 24.7R² = 0.92

0

20

40

60

80

100

120

15 65 115

Es

tim

ate

d H

eig

ht

(m)

Field Measured Height (m)

y = 0.61x2 -13.86x + 39.09

R² = 0.97

y = 0.39x2 + 6.18x + 80.66

R² = 0.94

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

0 100 200

Yie

ld (

kg

/ha

)

Plant Height (m)

FieldMeasured

PC Estimated

y = -0.23x2 + 57.95x - 3005R² = 0.43

y = -0.12x2 + 30.94x - 1276.6R² = 0.42

0

100

200

300

400

500

600

700

800

900

90.00 110.00 130.00 150.00

Lin

t Y

ield

(k

g/h

a)

Plant Height (cm)

Estimated Field Ruler-Measured

y = 7139x – 139R² = 0.74

0

100

200

300

400

500

600

700

800

900

0.06 0.08 0.1 0.12 0.14

Lin

t Y

ield

(kg

/ha)

CUC

• Glyphosate

• Dicamba

2018

2017

• 34 varieties

• 102 plots

• 3 reps

• Compared with military UAVs commercial/customer UAVs are much lower grade in product life, quality and reliability. So take a good care and be careful in flight operation. Don’t expect too high

• Payload – 10 - 50kg?

• Power - Electric UAVs dominate the market. Presently battery can only power the UAV flight duration 20-30 minutes for multirotors depending on payload and weather conditions. Fixed-wing may last longer (45-60 min) …

Limits of UAVs

• Bandwidth - Image cloud uploading, computing, downloading and visualization are still slow with practical requirements

• Image processing (mosaicking) may have system artifacts and smooth out the original information

• …

Limits of UAV Data Processing

• Technical Restrictions

• Batteries

•GPS/GNSS

• Law and Rules

• Visible range

• Speed limit

• Ceiling requirement

• Weight and size

UAV Law and RulesFly farther, faster, higher (Lower)

with more power?

• Safety, privacy and social management

• Safety

• Technical: Sense and avoid

• Laws and rules

US FAA

• Hobby or commercial?

• Register any aircraft if over 0.55 lbs (0.25 kg).

• Operators must pass FAA pilot examination and get FAA certified

• If less than 55 lbs (25 kg), certified operator can fly UAV under 400 ft (120 m) in the visible distance

• Certificate of Waiver or Authorization (COA)

Other countries and regions?

UAV Law and Rules

• International Organization for Standardization (ISO) working group (ISO/TC 023/SC 06/WG 25)

• Unmanned aircraft systems

• General specification

• Product systems

• Operation and procedures

…

• UAS Traffic Management

• Standards regarding UAV spraying systems

• American National Standards Institute (ANSI) Unmanned Aircraft Systems Standardization Collaborative (UASSC)

UAV Standards

8,800 lb (4,000 kg)

UAV Indoor Operation

Farm-Scale Remote Sensing Big Data

Service

Layer Level Block Size (o) Sphere

Dimension

(km)

Pixel Size (m) Scale

1 1 0.05 5.57 5.57 1:50,000

2 0.025 2.78 2.78 1:25,000

3 0.01 1.11 1.11 1:10,000

2 4 0.005 0.56 0.56 1:5,000

5 0.0025 0.28 0.28 1:2,500

6 0.001 0.11 0.11 1:1,000

3 7 0.0005 0.06 0.06 1:500

8 0.00025 0.03 0.03 1:250

9 0.0001 0.01 0.01 1:100

Artificial Intelligence and UAVs

• Smart flight control and navigation

• Automated to autonomous

Fly itself with AI by sensing environment and navigating without human input (DroNet, a lightweight residual convolutional neural network (CNN) architecture)

Remote Sensing Pattern Recognition

• Machine Learning/Soft Computing

• Deep Learning

• Generative adversarial networks (GANs) – Learn and create

Technologies for next-generation industry/agriculture

• Innovated methods, optimized algorithms,

massive data, and super computing power for

smart/intelligent industry/agriculture in cyber-

physical structure

• IoT (Internet of Things) • Automation

• Advanced materials, mechanical and

electronics science and technology

• Ubiquitous collaborative robots

• AI/DL to advance the perception of the

robotic systems

• Real-time process data processing, analysis, control and adaptation

UAV Education

• University, STEM (Science, Technology, Engineering and Math) and K12 curriculum

• Robotics competitions

UAVs (Drones) are getting More Capable and Smarter

amphibious

More user-friendly operational (for dummy)

Intelligent navigation (autonomous)

Safer

Environmentally adaptive

Standardized design,

manufacturing, configuration,

operation and service with

enhanced law and rules

through social and educational

management

THANK YOU!