1. Building A Brain With Raspberry Pi And Zulu Embedded JVM
Simon Ritter
Deputy CTO, Azul Systems

2. Introduction

3. Inspiration

4. Human Brain Neocortex

5. Hierarchical Temporal Memory
Theoretical framework for how the neocortex works
Models neurons and synapses and how they interact
Also includes dendrites and precise synapses
This is different to basic neural network topologies
HTM simulated neuron far more complex
More adaptable than AI neuron
links decay over time if not used (temporal)
Potentially much better at learning

6. Project Hardware

7. Raspberry Pi Fantastically successful development board
Cheap, powerful, great connectivity
Original boards $25, Pi Zero only $5!
Officially launched on Febuary 29th 2012
First production run was 10,000 boards
RS reported over 100,000 pre-orders in one day
Crashed servers
Over 10 million boards have been sold (Sept, 2016)

8. Raspberry Pi Variations
Model A
Model B
Pi Zero
ARM v6
700 MHz
ARMv6/7/8
700/900/1200 MHz
1 GHz
1 core
1 or 4 core
256 Mb
512 Mb
1 Gb
1 x USB
No Network
2/4 USB
100 Mb Ethernet
WiFi/Bluetooth (v3)
1 micro USB
No ethernet
WiFi/Bluetooth (Zero W)

9. Pi Brain Architecture Raspberry Pi Cluster Client Laptop
(Emulating HTM)
Network
Client
Laptop

10. Raspberry Pi Cluster (Mk 1)

11. Raspberry Pi Cluster (Mk 2)

12. Cluster Hat Raspberry Pi 3 Master Pi Zero W USB GPIO Slaves PORT 1
Alert LED
GPIO 21
Pi Zero W
Slaves
USB
GPIO

13. Java For Embedded
Applications

14. Why Java? “Write once, run anywhere”
Easy to move code between different platforms
Wide range of libraries and frameworks
Takes a lot of hard work out of app development
Simple interfacing
Pi4J and DIO libraries for Raspberry Pi
Simple concurrency
Comprehensive multi-threading support
Native code integration for complex situations
JNI

15. Azul Zulu Java For Embedded
Built from OpenJDK code base
Passes all TCK tests
Ports for Intel, ARM, PowerPC
ARM v6, v7, v8
Soft and hard float
64 bit close to release
C2 and C1+C2 compiler enhancements
Drop in replacement for other JVMs
No licensing restrictions (FoU)

16. Machine Learning In Java

17. Deep Learning Basics Neural network nodes
Neuron takes weighted inputs that can change
Organised as layers
Non-linear processing
Different levels of problem abstraction
Good analogy to HTM model
Unsupervised learning
No initial data set
Learn by experience

18. Neural Network Elements: Node1 W0 X1 W1 λ Σ W2 X2 W3 X3

19. Neural Network Elements: Layers
Output Layer
Hidden Layer
Input Layer

20. DeepLearning4J (DL4J)
Toolkit for building, training and deploying Neural Networks
provides tools to configure neural networks and build computation graphs
DataVec
Data ingestion, normalization and transformation into feature vectors
Arbiter
Search the hyperparameter space to find the best neural net configuration

21. DL4J Core Classes NeuralNetConfiguration.Builder backprop list
optimizationAlgo
MultiLayerConfiguration.Builder
tBPTTForwardLength/tBPTTBackwardLength
Layer (DenseLayer, OutputLayer)
nIn/nOut

22. Minecraft And Project Malmo

23. Minecraft Immensly popular ‘construction’ game
Especially amongst children
Different modes: creative, survival, hardcore
PC (first) version written in Java
121 million copies sold worldwide (all versions)
Probably most successful Java desktop app ever
Mojang aquired by Microsoft in 2014

24. Minecraft

25. Project Malmo Microsoft Labs project
AI experimentation platform built on top of Minecraft
Mod for Java version of Minecraft
Libraries to write AI agents to act as player in game
Including Java
Image courtesy of Microsoft

26. Malmo Example Code (1) AgentHost host = new AgentHost();
ClientPool clientPool = new ClientPool();
clientPool.add(new ClientInfo(" "));
MissionSpec mission = new MissionSpec();
mission.timeLimitInSeconds(120);
mission.requestVideo(640, 480);

27. Malmo Example Code (2) MissionRecordSpec missionRecord =
new MissionRecordSpec("./saved_data.tgz");
missionRecord.recordCommands();
missionRecord.recordMP4(20, );
missionRecord.recordRewards();
missionRecord.recordObservations();
host.startMission(mission, clientPool,
missionRecord, 0, "TEST");

28. Minecraft Interaction
MissionSpec
setModeToCreative
startAt/endAt
drawBlock/drawLine/drawSphere
AgentHost
startMission
peekWorldState
sendCommand

29. Putting It All Together

30. Demo Software Architecture
Pi Zero 1
(Input) TC
DL4J
Raspberry Pi 3
Demo Controller
Malmo Library
Pi Zero 2
(Fliter)
DL4J TC
TCP/IP
Pi Zero 3
(Fliter)
DL4J TC
Laptop
Minecraft with
Malmo mod
Pi Zero 4
(Output)
DL4J TC

31. Demo Scenario Original idea: Have agent build a house
Turns out that’s really hard
Demo revision 1
Have the agent build a wall
Simpler, but not much, than building a house
Demo revision 2
Have agent walk around without getting killed
Easier to develop a learning model

32. Applying ML To Minecraft
DL4J uses vectors to represent data
For our simple example use 3D vector
2D position and block type for next move
Start with random moves
Record success/failure (live/die)
Use reinforcement to (hopefully) learn how to navigate minecraft world safely

33. Conclusions & Future Directions

34. Conclusions Embedded hardware can be used for AI/ML/DL
Model on the structure of the human brain
Cheap to build multi-core, multi-CPU cluster
Java is an ideal language for this
Easy to learn
Simple threading model
Many, many useful libraries
Minecraft is a fun tool for AI testing and development

35. Future Directions More work on ML/DL for Minecraft
Make a better player!
Investigation of jCUDA
Use NVidia Jetson TK1 board
Use of machine vision
OpenCV Java bindings

36. Further Information www.raspberrypi.org clusterhat.com
deeplearning4j.org
minecraft.net
Code to be published on Github after some cleaning up
Will post a blog about this

37. Demo Time!