Rise of the Robots Australia’s Robotic Future Dr Airlie Chapman
Rise of the Robots
Australia’s Robotic Future
Dr Airlie Chapman
Melbourne Information, Decision and
Autonomous Systems (MIDAS)
Fundamental research areas:• Networked dynamical systems
• Computational engineering for dynamic
systems
• Human-centric autonomous systems
• Legal and societal implications of
autonomy
Large and complementary group of control, automation and optimisation
researchers (100pp) across multiple Departments in the University of Melbourne
(CIS, EE, ME, Science, Psychology)
Applications:• UAVs and UGVs
• Robotics
• Precision manufacturing
• Smart irrigation channels
• Powertrain control and calibration
• Power systems and microgrids
• Gas turbines
Research sponsors include:
Experimental Facilities
…with Energy:
Transient reciprocating
engine dynamometer
6-axis CNC with large work
piece at ANCA Motion
Network simulators
and rapid prototyping
control units
Motion capture
and robotic
platforms
Simulation capability
includes:
…with industry:
Vibration
control:
…with Flight:
Campus 2 autonomous
systems lab
…coming soon:
Irrigation network control
• Irrigation ~70% of all water use
• Smart wastegate control
– Modelling of irrigation channels
– Decentralised control
– A working “Internet of Things”
• Reported 20-30% water savings
• Recognised with 2008 ATSE Clunies Ross Award
Australia’s Robotic Landscape
• 18th for global automation by the International Federation of
Robotics
• 1st country to automate its ports
• Predicted to deliver $2.2 trillion dividend over the next 15 years
• 1100 companies support the robotic industry
as service business within major companies
or SMEs for niche markets
• Manufacturing robots accounts for 86%
robots (International Federation of Robotics)
• Drivers: Price, innovative applications,
consumer demand
Mobile Robotics
- Vehicles that act independently
Tesla – Autopilot
Roadmap for Mobile Robotics
• Motion control
• Navigation and mapping
• Sensors and predictions
• Emerging Technologies
• Challenges and a path forward
• Examine the current prediction of a system
• Design actions (control) to acquire a desired output
Motion Control
Courtesy: TechXplore
Cruise Control
Toyota Cruise Control
Precision
Volvo Dynamic Steering
Livestock
SwagBot – University of Sydney
Underwater
COTSbot – Queensland University of Technology
Transportation
Haulage train – Rio Tinto, Pilbara
• Plan paths and motions to navigate in its environment
• Assemble a map of the environment to plan over
Navigation and Mapping
Mining
Caterpillar – Haulage
Agriculture
John Deere – Tractor and Planter
Mine Stope Mapping
Emesent – CSIRO’s Data61
Infrastructure Maintenance
UTS Blasting Robot (ABC News)
• Sensors: Gives the ability to see, touch, and hear its
surroundings
• Prediction: Estimation of the state of the world, e.g., Computer
vision, AI, deep neural networks
Sensors and Predictions
Courtesy: Heriot-Watt/Audi
Sensors
Mobile Sensors
Autonomous Crop Interaction
Rippa – University of Sydney
UAV – Unmanned Aerial Vehicles
AgLoop TV
Aerial Monitoring and Inspection
AgLoop TV
Horteye – University of Melbourne
Multicopters Inspection
• Delivery and placement
• One to Many
• Collaborative robotics
Emerging
Delivery and Placement
Project Wing – Alphabet (WSJ)
Multi-Package Delivery
MIDAS Air – University of Melbourne
One to Many - Cooperative Robotics
Kiva Systems – Amazon Robotics
Cooperative Warehousing
Collaborative Construction
Digital Fabrication – ETH Zurich
Collaborative Coverage
MIDAS Networks – University of Melbourne
Challenges and a Path Forward
• Technical challenges: GPS denied environments, cluttered
spaces, battery capacity, communication and computation
limitations
• Promote government frameworks for autonomous systems
• Establish test sites to trial technologies
• Develop robotic clusters of innovation with strong industry links
• Publicity: Safety, accuracy and productivity
• Success: Robots to tool
Thanks.