ロボットとのコミュニケーションマニュアル / Manual for Communicating with Robots

2025年5月3日(土・祝)〜6日(火・祝)、早稲田大学にて開催されたアカデミーキャンプ 2025GWeeeeeeK「がんばれ!!ロボコン」で使用したスライドです。

Transcript

1. 2025GWeeeeeeK !! 2025GWeeeeeeK !! — 2025-05-03 – p.1/20
2. https://speakerdeck.com/ks91/collections/academy-camp 2025GWeeeeeeK !! — 2025-05-03 – p.2/20
3. ϩϘοτ ਓؒ 
   -JHIU ϩϘοτ HIU U ਓؒ 
   ·ͩ  େ೴ αʔόʔ খ೴ Ի੠Ͱ΍ΓͱΓ ΫϥΠΞϯτ ͭͳ͗ Ϟʔλʔ΍ηϯαʔ جຊ ࢦྩ ͍·ɺ͋ͳͨͷ೴ʹ ௚઀࿩͔͚͍ͯ͠·͢ ϩ Ϙ ỽ τ ͷ छ ྨ ͝ ͱ ʹ ҟ ͳ Δ HQUPSFBMUJNFQSFWJFX HQUPNJOJ 01 OpenAI GPT-4o-realtime-preview GPT-4o-mini 2025GWeeeeeeK !! — 2025-05-03 – p.3/20
4. — AI Freenove Raspberry Pi 4B (4B OK) https://www.amazon.co.jp/dp/B08M5DXS2P https://www.amazon.co.jp/dp/B0BNDQFRP1 ( ) https://github.com/ks91/Freenove_Hexapod.git https://github.com/ks91/Freenove_Tank.git https://github.com/ks91/01 ( : )( git clone ) RasTech Raspberry Pi 4 Model B (RAM 8GB) (microSD ) https://www.amazon.co.jp/gp/product/B0C8BBVVXK 18650 ( )( ) : https://www.amazon.co.jp/dp/B0992G1LB9 ( ( ) 4 ) : https://item.rakuten.co.jp/oremeca/10000911/ (2 ) SSH (macOS Windows (10 )) (iPad Termius ) RealVNC Viewer https://www.realvnc.com/en/connect/download/viewer/ API ( ) ChatGPT Plus ( GPT ) 2025GWeeeeeeK !! — 2025-05-03 – p.4/20
5. : 01 Light Wi-Fi ( ; Wi-Fi ) iOS / iPadOS https://apps.apple.com/jp/app/01-light/id6601937732 android https://play.google.com/store/apps/details?id=com.interpreter.app&hl=ja 01 Light . . . “Continue ( )” “I have backed up all my files ( )” “I understand the safety implications of running AI generated code on my computer (AI )” “I have read and agreed to the 01 App terms of service (01 )” “Tap to scan ( )” 2025GWeeeeeeK !! — 2025-05-03 – p.5/20
6. ON/OFF ON (0) (1) CTRL/POWER LOAD CTRL/POWER LOAD (2) iPad (mini) (3) iPad RVNC Viewer OFF ( ) (1) (2) (3) Shutdown (4) LED LOAD → CTRL/POWER micro SD ( ) 2025GWeeeeeeK !! — 2025-05-03 – p.6/20
7. (0) iPad RVNC Viewer (1) >_ iPad (2) ( ) run.sh $ cd 01/software $ ./run.sh ↑ QR 2025GWeeeeeeK !! — 2025-05-03 – p.7/20
8. run.sh lxterminal --command ’cd ~/Free*/Code/Server; sudo python main.py; bash’ sleep 2 lxterminal --command ’cd ~/Free*/Code/Client; python Main2.py; bash’ sleep 2 poetry run 01 --server livekit --qr --multimodal 1 : main.py ( ; ) 3 : Main2.py ( ; ) 5 : 01 ( ) 2025GWeeeeeeK !! — 2025-05-03 – p.8/20
9. 01 Light “Tap to scan” ( ) QR QR INFO DEBUG WARNING control+C run.sh 5 cd 01/software ./run.sh / ( ) . . . iPad 2025GWeeeeeeK !! — 2025-05-03 – p.9/20
10. (1/2) (1) $ cd 01/software/source/server/livekit $ nano multimodal.py nano (2) voice : model = openai.realtime.RealtimeModel( instructions=instructions, voice="sage", # {"alloy" | "shimmer" | "echo" | "ash" | "ballad" | "coral"... temperature=0.6, : 2025GWeeeeeeK !! — 2025-05-03 – p.10/20
11. (2/2) (3) voice="shimmer" voice="sage" “alloy” → “shimmer” → ( ) “echo” → “ash” → “ballad” → “coral” → “sage” → ( ) “verse” → (4) control + O 01 01 Light control + X nano 2025GWeeeeeeK !! — 2025-05-03 – p.11/20
12. : $ cp *hexa*.py multimodal.py : $ cp *tank.py multimodal.py cp * ( ) ← multimodal.py 01 multimodal-{hexapod|tank}.py multimodal.py $ git checkout 2025GWeeeeeeK !! — 2025-05-03 – p.12/20
13. Python OK “01/software” Python (sftp ) 2025GWeeeeeeK !! — 2025-05-03 – p.13/20
14. — 2025GWeeeeeeK !! — 2025-05-03 – p.14/20
15. (1/3) ON/OFF (OFF ) ( ) ( ) / (2∼10; 8) / ( ) / / x 1 (x y 0 ) : (1) 3 (2) 1 ( ) 2025GWeeeeeeK !! — 2025-05-03 – p.15/20
16. (2/3) / (90 : , 180 : , 60 : ) (90 : , 0 : , 180 : ) ( ) ( : -40∼40, : -20∼20) ( ) ( ( ) / ( ) / ( ) : -15∼15) ON/OFF (ON ) / ( ) LED ( : 1∼5) LED ( : 0∼255) (200 ) 2025GWeeeeeeK !! — 2025-05-03 – p.16/20
17. (3/3) opencv ∼ ( : cm) 2025GWeeeeeeK !! — 2025-05-03 – p.17/20
18. — 2025GWeeeeeeK !! — 2025-05-03 – p.18/20
19. (1/2) ( ) / ( ) (700∼2000) 2000 700 / (90 : , 150 : ) / (90 : , 150 : ) LED ( : 1∼5) LED ( : 0∼255) (200 ) 2025GWeeeeeeK !! — 2025-05-03 – p.19/20
20. (2/2) opencv ∼ ( : cm) ( 1 ( ) ) ^^; 2025GWeeeeeeK !! — 2025-05-03 – p.20/20