Transcript Outline of Approach - University of Colorado Boulder

Autonomous Sensor and Control
Platform Rover
•
•
•
•
•
Tae Lee
Josh Reitsema
Scott Zhong
Mike Chao
Mark Winter
Brief Description
• An easily extensible autonomous rover
platform.
• Initially equipped with SONAR, IR, and
bumper sensors.
• A camera for remote viewing and control.
• Functionally extendable control program
makes the rover extremely versatile.
Goals
• Central processing and control circuitry.
• Control for devices: Motors, SONAR,
IR, and bumpers.
• Communication with a computer via
serial or possibly WIFI for programming
and remote control.
• Camera mounted on rotating platform,
to provide computer controllers with a
video display
Extensions
• WIFI communication with computer.
• DSP for video to heighten the rover’s
“senses”
Cost and Returns
• Initial costs are low because of the
availability of a basic rover platform for
prototyping, equipped with motors and
sensors.
• Given the high versatility of the
Autonomous Sensor and
Control(ASAC) platform, we can expect
high demand from companies requiring
an autonomous unit for sensor data
collection in inaccessible areas.
Outline of Approach
Autonomous Control
Behavior Software
Sensor Drivers and Hardware
Communication Software
Communication Hardware
Low Level Control Hardware
Needed Hardware Subsystems
• Main Controller Board
• Sensor Interface
Hardware
• Motor Controller
Hardware
• Communication to
Human Interface
User Control
Main Board
Robotic
Movement
Sensor Inputs
Main Controller Board
• Based on 2120 board
• Motorola 68HC11 processor, Xlinix logic
chip, multiple I/O buslines for input to
sensors and output to motors
• Needs to be programmable in C, C
compilier
Other Hardware Systems
• Each sensor, video, IR, wheel encoders,
etc., needs to be interfaced into the main
controller board
• Communication will be wi-fi and direct
connect
• The motors must be set precisely with
correct current and voltage
Software Subsystems
• Drivers for hardware
will be written in
Assembly
• High level control, i.e.
decisions and
behavior, will be
written in C
• Control will be
accessed through a
graphical user
interface
Implementation of Various
Subsystems
•
•
•
•
•
Chassis
Mainboard
Power
Communications
Sensors
Chassis
• Provided with a stripped down vehicle
– Frame
– 2 X 12V stepper motors
• Controlled using a Generic H-Bridge.
– Sonar sockets placed around the cart
• Possibly implement if time permits.
Mainboard
• Using a 68HC11 to process commands
• FPGA
– Will handle expanded IO to sensors.
– Will execute movement commands.
Power
• Regulators will provide 3 different
voltages.
• 12V
– Motors
• 5V
– Motorola 68HC11
– Xilinx FPGA
• 3.3V
– Communication Module
Communication
• Programming handled through a serial cable
• Integrate a WiFi module
– Lantronix WiPort.
• Otherwise fall back to RC or wired
communication.
Sensors
• Camera will feed data to a monitor
– Time permitting data will be transmitted over
wireless interface.
– Otherwise it will have its own independent
transmission (baby monitor).
• Hopeful expansion
– Integrate existing sonar ports.
Division of labor and
responsibilities
Scott
– Main board and processor
– Specification of interface between each different
module (WiFi, Sonar, etc)
Tae
- Main board and processor
- Manual editor
- Ultrasonic Sensors
Division of labor and
responsibilities
Mike
–
–
–
WiFi Interface
Control of Rover
Sensor Data
Josh
–
–
–
–
Video Camera Interface and Control
Movement Circuit
Ultrasonic Sensors
High Level Control Software
Division of labor and
responsibilities
Mark
–
–
–
–
Test and Document Current Hardware
Drivers
Data Processing Algorithms
Video Interface
Division of labor and
responsibilities
Everyone
–
–
–
–
Documentation
User manual
Tech Support
Trips to get food
Considerations
• Project to Complex?
• Too many peripheral sensors
• Do we have enough time to do all this
• Is there enough processing power
• PCB
• Eliminate Noise
• Wire wrap problems go away
• Ultrasonic Sensors
• How many do we need for accuracy
• How are we going to connect them
• Video Feed
• Stand alone unit with no integration
• Wi-fi using serial
• Is it fast enough
Risks
• Do we have what it takes
• Knowledge of components
• Can we learn really, really fast?
• Unexpected Complexities
• Bad parts, “what's that smell” during testing
• How long to fix the unexpected?
• Complications during integration of
modules
• Use of old motors
Contingency Plan
• Make sensor sub-systems very modular
• Leave out the flash until the end
• Make sure the basics work first
• Wired communications
• Back up for wi-fi
• Order extras if we can
• Replace old motors with newer model
• Data sheets, Support
Thanks
• Questions?