WOW, November’s dinner was very info-packed! A Huge Thank You to our speakers!
Here are some brief takeaways:
- Clearpath Robotics aims to be the leader in self-driven vehicles. They build industrial robots for mining, defense, and agriculture.
- There are eight core competencies of robotics: Systems Design, Perception, Mapping, Navigation, Control and Simulation. In Catherine Holloway’s Test Technician role, she heavily tests the robots in these various competencies. More information about the core competencies can be found here: http://www.clearpathrobotics.com/services/
- Some of the programs and languages used by Clearpath are: Robot Operating Systems, Gazebo, Raspberry Pi, and Python.
- Three cool demos:
- There is heavy testing between human behaviours and robot interactions. Here is the video “Kids Abusing Robot”:https://www.youtube.com/watch?v=pnSLGcoiFqg
- To create your own robot world, you could use a program calledGazebo. You could also control one through a Nintendo Wiimote and Raspberry Pi Robot.
- Demo about Error Analysis; a comparison of how much the robot moved in real life and how much it thought it moved.
- In Teyvonia’s role as a Perception Engineer, she makes sure the robots are well-calibrated to sensors. Some of those sensors are: visual slam, visual odometry, keyframe selection, mapping, loop closure detection and place recognition.
- Some challenges in the sensors are: weather, lighting, shadows, seasons, non-discriminative regions, regions, viewpoint changes, etc.
- Coordinating tasks with multiple robots is more challenging.
Question and Answer period:
- Do robots have temperature sensing features?
- Yes, some of them do detect adverse heating or cooling.
- Have beacons been considered for mapping purposes?
- Visual slam is still highly experimental and there are many challenges. There are already robots in factories that drive around using bar codes or magnetic tapes; disadvantage is that it requires infrastructure.
- What are the technological roadblocks?
- Interaction with real human beings, e.g. factory workers, not only robotics engineers who are deferential to robots.
- Ethical debates about building potentially dangerous robots. Educating the public about what the company is trying to do, social issues, e.g. job security in community.
- How do you feel about drones doing deliveries?
- It would be very, very noisy. It would need faster and bigger motors. It would be very annoying and everyone would probably need to wear noise-cancelling headphones.
- How close are we to the possibility of iRobot?
- Everything is already there; the possibilities are endless, we just need to do it.
- Where is the Robotics field going?
- Research is moving fast and there are lots of opportunities here. Robots should be capable, but not “too” intelligent. In Robotapocalypse talks, we project ourselves to robots. Robots have different needs than humans but we always imagine them to behave like one where they will go revenge mode. Having said that, they can be programmed to behave as humans by displaying gestures.