Deep Learning Robot Demo - ROS and Robotic Software Makespace, - - PowerPoint PPT Presentation

▶
deep learning robot demo ros and robotic software
SMART_READER_LITE
LIVE PREVIEW

Deep Learning Robot Demo - ROS and Robotic Software Makespace, - - PowerPoint PPT Presentation

Aug 04, 2023 264 likes •703 views

Deep Learning Robot Demo - ROS and Robotic Software Makespace, Cambridge UK 22nd February 2016 About Me Games VR Webisodes / Entertainment Software development Startups simon@robotlux.com @eurodemanding Sold Sold

slide-1
SLIDE 1

Deep Learning Robot Demo - ROS and Robotic Software

Makespace, Cambridge UK 22nd February 2016

slide-2
SLIDE 2

About Me

  • Games
  • VR
  • Webisodes / Entertainment
  • Software development
  • Startups

Sold ✔ Sold ✔ simon@robotlux.com @eurodemanding

slide-3
SLIDE 3

Insect Bot

slide-4
SLIDE 4

Teach Bot

slide-5
SLIDE 5

Pi Camera Bot

Pi Camera Bot

slide-6
SLIDE 6

SLAM

  • Simultaneous Localization and Mapping
  • Localization: How does a robot know where it is

in a world of untrustworthy sensors?

  • Mapping: How can it make

a map when it doesn’t know where it is?

slide-7
SLIDE 7

Life’s too short. What can we steal?

slide-8
SLIDE 8

Robot Operating System

  • Not just for robots
  • Not an operating system

ROS = An open source framework and a collection of packages that are useful in robotics

slide-9
SLIDE 9

ROS Packages

  • Navigation: SLAM, autonomous navigation…
  • Robot Arm: Kinematics, inverse kinematics…
  • Hardware Drivers: LIDARs, sound, motors, vision…
  • Interfaces: OpenCV, Caffe, Speech to text…
  • …
slide-10
SLIDE 10

The History of ROS

slide-11
SLIDE 11
slide-12
SLIDE 12
slide-13
SLIDE 13

Parrot AR.Drone 2.0 Elite with ROS drivers

slide-14
SLIDE 14

Erle Robotics

slide-15
SLIDE 15
slide-16
SLIDE 16

ROS Architecture: Nodes and Topics

Node – independent software process that publishes and subscribes to Topics Topic – A stream of structured data messages

LIDAR

Laser Range Finder

SLAM Object Recognition Safety Override

slide-17
SLIDE 17

“washing machine”

  • bjects_in_view

greeter “hello, washing machine” text_to_speak

greeter.py

slide-18
SLIDE 18

ROS Navigation Stack

Enough to do SLAM and autonomous navigation

slide-19
SLIDE 19

Where in the stack do you want to experiment?

Behaviours Faculties Hardware Drivers Electronics Mechanics Grippers, wheels, legs, chassis… Microcontrollers, IMUs, sensors… Arduino code / C++ SLAM, navigation, object recognition… Play with the dog

slide-20
SLIDE 20

The Deep Learning Robot

www.autonomous.ai $1000 = GBP 700

slide-21
SLIDE 21

Kobuki Mobile Base

  • 2 wheel, differential

drive

  • Wheel encoders
  • 3 bump sensors
  • 1 cliff sensor
  • Wheeldrop sensor
  • Gyroscope
  • IR-based docking
  • USB communication

with robot motherboard

slide-22
SLIDE 22

nVidia Jetson TK1

Robot motherboard:

  • ARM CPU
  • 2 Gb RAM
  • 16Gb Flash
  • nVidia GPU with

192 CUDA cores

  • Wifi & Bluetooth

Principal value add is CUDA acceleration

  • f deep learning

tools

slide-23
SLIDE 23

Asus Xtion Pro Live

  • Camera with RGBD

(RGB + depth output)

  • Uses infrared to

rangefind

  • Microphone
  • USB communication

with Robot motherboard

  • Primesense, succesor

to Kinect

  • Intel RealSense3D is

like succesor to this

slide-24
SLIDE 24

Demo

slide-25
SLIDE 25

Great, free, introductory course on the maths of SLAM, autonomous navigation

https://www.udacity.com/course/artificial-intelligence-for-robotics--cs373 Artificial Intelligence for Robotics UDACITY

slide-26
SLIDE 26

Deep Learning Neural Networks Machine Learning

If X has features a, b, c, d… then what is Y ?

If X is age 42 then what is their net worth ? If X is an email with words “viagra”, “cheap”… then is it spam ? If X is a house with 3 BDR, centre of Cambridge and in lousy condition then what is the price ? If X is an image with pixels (1, 2, 3…10,000) then is it showing my grandmother?

slide-27
SLIDE 27

Neural Networks

a b c d e f

X Y

Back Propagation Pixels “Grandmother”

slide-28
SLIDE 28

Neural Networks

  • Retro and futuristic
  • They work now (but didn’t in the 80s) because of

– Fast CPUs – Fast GPUs (all thanks to gamers) – Large datasets

  • Deep Learning
  • CNN: Convolutional Neural Networks
  • RNN: Recurrent Neural Networks
  • ….
  • Is back propagation the fundamental computational

building block of the human brain?

slide-29
SLIDE 29

Caffe

  • Tool for designing, training and testing neural networks, especially

related to vision

  • CUDA accelerated
  • Widely used in research
  • Pre-installed on the robot (along with similar Google TensorFlow,

Theano etc.)

slide-30
SLIDE 30

Demo

slide-31
SLIDE 31

Survey

slide-32
SLIDE 32

What next?

slide-33
SLIDE 33
slide-34
SLIDE 34
slide-35
SLIDE 35

10 print “piss off” 20 goto 10

slide-36
SLIDE 36
slide-37
SLIDE 37
slide-38
SLIDE 38

10 print “piss off” 20 goto 10

slide-39
SLIDE 39
slide-40
SLIDE 40
slide-41
SLIDE 41
slide-42
SLIDE 42
slide-43
SLIDE 43

Thanks

simon@eurodemand.com www.artificialhumancompanions.com @eurodemanding