A Robotic Unmanned Aerial Vehicle for Environmental Research and Monitoring

1
A Robotic Unmanned Aerial Vehicle for
Environmental Research and Monitoring

• Alberto Elfes, FAW, Germany
• Mario F. M. Campos, UFMG, Brazil
• Marcel Bergerman, CTI, Brazil
• Samuel S. Bueno, CTI, Brazil
• Gregg W. Podnar, CMU, USA
• Contactelfes_at_faw.uni-ulm.demarcel_at_ia.cti.brssb
  ueno_at_ia.cti.br

2
Environmental and Biodiversity Research and
Monitoring

• Significance
• Environment protection
• Sustainable use of natural resources
• Monitoring climate change
• Protection of living species and natural habitats
• Survival of the human race
• Difficulty requires massive amounts of
  information
• atmosphere
• oceans
• land masses
• biosphere

3
Environmental Data Acquisition

• Major difficulties faced by developing countries
• large size
• limited transportation infrastructure
• cost of data acquisition sensor networks,
  transportation, field missions, risks, etc.
• Limitations of current data acquisition sources
• satellites limited spatial resolution,
  predefined spectral bands, limited geographic and
  temporal sampling, limited variables that can be
  measured
• manned flights interferes with the environment
  (noise, turbulence), costly, often of short
  duration
• vessels special weather requirements, costly
• sensor networks complex and costly, difficult to
  deploy and maintain, limited coverage
• field expeditions logistics, cost, risk,
  difficulties associated to reaching and leaving
  the field

4
Robotic Unmanned Aerial Vehicles (RUAVs)

• Adding robotic capabilities to UAVs can
  significantly broaden their utilization
• Combination of navigation and mission sensors
  allows sensor data to be registered to time and
  positional information for analysis purposes
• Data gathering missions and multi-modal sensing
  at high resolution based on
• low flying altitude
• time windows of observations can be defined for
  each individual mission
• flexibility in defining the geographical regions
  to be observed and how they will be covered

5
RUAVs as Autonomous AerialSensing Platforms

• RUAVs can operate as independent autonomous
  aerial platforms carrying a complement of sensors
  onboard
• Larger RUAVs could carry also a miniaturized
  Flying Laboratory, or FLAB, composed of
• sample analysis stations
• computer-controlled microscopes
• micromanipulators
• sensors
• microcameras (operation and supervision)
• high-speed LAN connection
• Internet connection

6
RUAVs for Sensor Probe Deployment

• RUAVs may operate as sensor package delivery
  systems deploying sensor pods in remote or
  inaccessible sites
• Low altitude, low speed vehicles are ideal for
  deployment and recovery of probes on the ground,
  rivers, lakes, oceans, and canopies
• Samples obtained may be returned to a base
  station of processed in situ by the FLAB
• Probe deployment and recovery would be done under
  the supervision of a team of field experts across
  the world

7
RUAVs for Deployment of Robots

• RUAVSs can be conceived as deployment platforms
  for ground or water exploration robots
• Robots may consist of sensors and other probes
  for detailed exploration
• A high level of autonomy will allow an
  exploration robot to perform its task
  unsupervised, and later rendezvous with the RUAV
  to return to a base station

8
RUAVs as Members of Exploration Teams

• RUAVs can be thought of as members of a team
  structure with other robots and humans operating
  in the field
• A properly selected and coordinated team of
  heterogeneous robots will greatly extend the
  capability of human beings to perform complex
  data acquisition tasks
• Cooperative inspection and handling applications
  can benefit from a partition of responsibilities,
  where the RUAV provides broad visual coverage and
  perception, and other robots execute close-up
  inspection and manipulation

9
Project AURORA

• Types of UAVs
• airplanes, helicopters, airships
• Airships are highly suited for environmental and
  biodiversity applications
• hovering capability
• stable, low noise and turbulence platform
• AURORA Autonomous Unmanned Remote Monitoring
  Robotic Airship
• Focus sensing, inference, control, and
  navigation technologies required for
  substantially autonomous robotic airships

10
AURORA I

• Physical characteristics
• length 9m
• volume 26 m3
• payload 10kg
• two engines and four control surfaces

11
AURORA I

• Onboard infrastructure
• PC104-based Pentium 133 MHz
• Real time Linux
• GPS receiver
• inertial navigation sensors
• radio link to ground station
• camera and video link
• Ground infrastructure
• PC Pentium 300 MHz
• Real time Linux
• DGPS receiver
• radio link to onboard station
• video link

12
Autonomous Navigation
13
Experiment on Vegetation Identification and
Tracking
14
Experiment on Cattle Identification and
Segmentation
15
Experiment on Perceptual Cooperation Among
Ensembles of Robot Agents
16
Towards Data on Demand

• The concepts discussed here can become part of a
  larger effort to provide environmental and
  biodiversity data on demand
• How? With a network of environmental databases
  queried by inference engines to retrieve already
  available data and plan the acquisition on new
  data not in the databases
• Planning the acquisition of new data
• sattelite imagery
• manned flight
• human expedition
• robotic team
• combination of some or all of the above