Rishabh Garg, Tesla Optimus — Challenges in High Performance Robotics Systems
A robot's behavior is influenced by the control policy, the software configuration, and electrical characteristics of the communication protocol.
When unexpected behaviors arise, it is not straightforward to root cause them to the RL policy, electrical characteristics, mechanical characteristics. This talk walks through some of these issues and explains what might cause the observed behavior.
We will talk about concrete issues that audience will be able to take away from and develop their understanding of physical systems. It will build intuition for what kind of issues to expect when commun…
Watch on YouTube ↗
(saves to browser)
Chapters (20)
Introduction to high-performance robotics challenges
0:15
The problem of unexplained robot behavior
0:54
Root cause analysis: policy vs. software
1:17
Designing a toy robotics system for analysis
1:24
System architecture: sensors, CPU, GPU, actuators, CAN bus
1:57
The initial, simple code loop
2:14
Expectation vs. reality: unexpected loop execution gaps
2:42
The impact of CAN bus data rate on loop execution
3:13
Potential solutions: accepting delay vs. multithreading
4:00
A new, pipelined design for reduced cycle time
4:32
New problems: "stuttering" and abnormal motor behavior
4:49
Data collection with external transceivers and "candump"
5:24
Expected vs. actual message plots: missed messages and jitter
6:12
Using cycle time plots to identify desynchronization
6:58
Transmit phase desynchronization: missed and queued data
8:03
Receive phase desynchronization: stale data and overcompensation
8:38
Resolving synchronization issues: kernel primitives and padding
9:25
The impact of logging on system performance
11:09
Reception and priority inversion
12:02
Conclusion and summary of key takeaways
Playlist
Uploads from AI Engineer · AI Engineer · 0 of 60
← Previous
Next →
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AI Engineer Summit 2023 — DAY 1 Livestream
AI Engineer
AI Engineer Summit 2023 — DAY 2 Livestream
AI Engineer
Principles for Prompt Engineering - Karina Nguyen (Claude Instant @ Anthropic)
AI Engineer
Announcing the AI Engineer Network: Benjamin Dunphy
AI Engineer
The 1,000x AI Engineer: Swyx
AI Engineer
Building AI For All: Amjad Masad & Michele Catasta
AI Engineer
The Future of Work: Toran Bruce Richards, Silen Naihin et al
AI Engineer
The Age of the Agent: Flo Crivello
AI Engineer
See, Hear, Speak, Draw: Logan Kilpatrick & Simón Fishman
AI Engineer
Building Context-Aware Reasoning Applications with LangChain and LangSmith: Harrison Chase
AI Engineer
Pydantic is all you need: Jason Liu
AI Engineer
Building Blocks for LLM Systems & Products: Eugene Yan
AI Engineer
The Intelligent Interface: Sam Whitmore & Jason Yuan of New Computer
AI Engineer
Climbing the Ladder of Abstraction: Amelia Wattenberger
AI Engineer
Supabase Vector: The Postgres Vector database: Paul Copplestone
AI Engineer
[Workshop] AI Engineering 101
AI Engineer
The Hidden Life of Embeddings: Linus Lee
AI Engineer
[Workshop] AI Engineering 201: Inference
AI Engineer
The AI Pivot: With Chris White of Prefect & Bryan Bischof of Hex
AI Engineer
The AI Evolution: Mario Rodriguez, GitHub
AI Engineer
Move Fast Break Nothing: Dedy Kredo
AI Engineer
AI Engineering 201: The Rest of the Owl
AI Engineer
Building Reactive AI Apps: Matt Welsh
AI Engineer
Pragmatic AI with TypeChat: Daniel Rosenwasser
AI Engineer
Domain adaptation and fine-tuning for domain-specific LLMs: Abi Aryan
AI Engineer
Retrieval Augmented Generation in the Wild: Anton Troynikov
AI Engineer
Building Production-Ready RAG Applications: Jerry Liu
AI Engineer
120k players in a week: Lessons from the first viral CLIP app: Joseph Nelson
AI Engineer
The Weekend AI Engineer: Hassan El Mghari
AI Engineer
Harnessing the Power of LLMs Locally: Mithun Hunsur
AI Engineer
Trust, but Verify: Shreya Rajpal
AI Engineer
Open Questions for AI Engineering: Simon Willison
AI Engineer
Storyteller: Building Multi-modal Apps with TS & ModelFusion - Lars Grammel, PhD
AI Engineer
GPT Web App Generator - 10,000 apps created in a month: Matija Sosic
AI Engineer
Using AI to Build an Infinite Game: Jeff Schomay
AI Engineer
How to Become an AI Engineer from a Fullstack Background - Reid Mayo
AI Engineer
The Code AI Maturity Model and What It Means For You: Ado Kukic
AI Engineer
AI Engineer World’s Fair 2024 - Open Models track
AI Engineer
AI Engineer World’s Fair 2024 — Keynotes & CodeGen Track
AI Engineer
AI Engineer World’s Fair 2024 - Open Models track
AI Engineer
AI Engineer World’s Fair 2024 — GPUs & Inference Track
AI Engineer
AI Engineer World’s Fair 2024 - Keynotes & Multimodality track
AI Engineer
From Text to Vision to Voice Exploring Multimodality with Open AI: Romain Huet
AI Engineer
The Making of Devin by Cognition AI: Scott Wu
AI Engineer
The Future of Knowledge Assistants: Jerry Liu
AI Engineer
Llamafile: bringing AI to the masses with fast CPU inference: Stephen Hood and Justine Tunney
AI Engineer
Open Challenges for AI Engineering: Simon Willison
AI Engineer
Lessons From A Year Building With LLMs
AI Engineer
From Software Developer to AI Engineer: Antje Barth
AI Engineer
Unlocking Developer Productivity across CPU and GPU with MAX: Chris Lattner
AI Engineer
Copilots Everywhere: Thomas Dohmke and Eugene Yan
AI Engineer
Fixing bugs in Gemma, Llama, & Phi 3: Daniel Han
AI Engineer
Low Level Technicals of LLMs: Daniel Han
AI Engineer
Emergence Launch: AI Agents and the future enterprise: Dr. Satya Nitta
AI Engineer
How Codeium Breaks Through the Ceiling for Retrieval: Kevin Hou
AI Engineer
What's new from Anthropic and what's next: Alex Albert
AI Engineer
Using agents to build an agent company: Joao Moura
AI Engineer
Decoding the Decoder LLM without de code: Ishan Anand
AI Engineer
Running AI Application in Minutes w/ AI Templates: Gabriela de Queiroz, Pamela Fox, Harald Kirschner
AI Engineer
Building with Anthropic Claude: Prompt Workshop with Zack Witten
AI Engineer
DeepCamp AI