Chao Cao (Co-Founder & CEO, Sancho - CMU robotics PhD, autonomy lead for CMU's DARPA Subterranean Challenge entries, ex-Boston Dynamics AI Institute) makes the case that the dominant approach to robot intelligence, pixel-based VLA models trained on massive datasets, is an expensive detour. Sancho bets on 3D geometric world models and test-time reasoning instead, and runs the entire stack on a single onboard compute module.

We cover the lasagna analogy for factory workflows, the math on the most expensive labor doing the lowest-value work, what five nines of reliability does to the data question, Waymo vs. Tesla as a data-quality argument, three years of sending robots into tunnels and caves, the October-to-March sprint from incorporation to NVIDIA's GTC keynote, the oversubscribed seed round co-led by Fusion Fund and Catapult, and why regulated cleanrooms are an easier place to start than a grocery store.

Chapters:

00:00 Intro

01:10 Roadmap and why the company is called Sancho

02:15 The lasagna problem: the gap between machines

04:51 $300K scientists loading machines by hand

06:04 Why fixed automation isn't the answer

07:02 Bet #1: 3D geometry over pixels

09:46 Bet #2: test-time reasoning over data scaling

11:25 Running the whole stack on onboard compute

13:11 Three years in the dark: DARPA SubT lessons

16:15 October to GTC in five months

18:35 The oversubscribed seed round

19:01 Two founders, two autonomy worlds

21:08 Why start in the most regulated environments

23:48 Hot takes: Waymo vs. Tesla and wasted data

25:58 What Chao wishes he knew before starting

27:30 Hiring: perception and loco-manipulation

29:50 Where to find Sancho

Learn more about Sancho: https://www.sancho.com

The OPTIM Update covers real-world AI, automation, robotics, industrial systems and AI Infrastructure for founders, investors, and operators. Hosted by Bogdan Cristei of Optim VC.

Subscribe to the newsletter: https://www.optim.vc

Follow Bogdan on LinkedIn: https://www.linkedin.com/in/bogdancristei/

We cover:
00:00 - Intro
00:46 - From McKinsey Bangkok to an MIT PhD with a camera on his head
02:36 - The Christmas pivot: from training humans to training robots
03:41 - Why robotics doesn't have its "internet" yet
04:50 - What the data product actually looks like
05:27 - Egocentric, tactile, teleoperation - every lab wants something different
07:12 - A factory capture from start to finish
07:53 - "My dad runs a factory" - the first test case
08:31 - Getting factory owners to trust you
09:40 - Industrial influencers and the factory network
10:12 - Scaling across India, Thailand, Vietnam, Indonesia
11:30 - What frontier labs learned from the pilots
12:57 - The gap between what labs want and what you can deliver
13:16 - Closing a $6M round and doubling the team in two weeks
14:00 - New verticals beyond manufacturing
14:42 - The synthetic data question
15:14 - Chaos theory and why sim data compounds errors
17:02 - Where defensibility lives in this business
17:20 - 80% of raw footage is unusable
19:29 - What the market gets wrong about robotics timelines
21:28 - Hot take: convergence to 2-3 big players
22:32 - The 10-year vision: becoming the Siemens of robotics
23:18 - Advice for robotics founders

Learn more about Vision Lab: https://thevisionlab.ai

The OPTIM Update covers real-world AI, automation, robotics, industrial systems and AI Infrastructure for founders, investors, and operators.

Subscribe: https://www.optim.vc

We cover:
00:00 - Intro
00:45 - What physical AI teams actually go through before training a model
03:16 - Why the existing robot stack was built for a different era
04:10 - What it's actually like setting up an SO-100 at home
05:21 - The leap from toy arms to real payloads
08:01 - What you get on day one with an Anvil dev kit
09:12 - What kilohertz-rate sensor fusion actually unlocks
11:19 - The false tradeoff between payload and force compliance
14:35 - Why vision alone isn't enough: the dentist analogy
16:15 - The economics of a $5,000 arm
20:01 - Scaling from 150 robots to 200 a month
21:30 - Why all customers came inbound
22:10 - Retention and repeat orders
24:47 - If open source isn't the moat, what is?
28:15 - Why the supply chain is a relationship, not a transaction
28:36 - How to do customization without becoming a services company
31:30 - How many of 1,500 new robotics startups survive 24 months?
34:52 - The most technically wrong thing teams are doing in 2026
38:57 - What happens when your whole fleet breaks and you don't know why
40:40 - What will look obvious in five years
42:29 - Where to learn more about Anvil

Anvil Robotics: https://anvil.bot

The OPTIM Update covers real-world AI, automation, robotics, industrial systems and AI Infrastructure for founders, investors, and operators.

Subscribe: https://www.optim.vc

We cover:
00:00 - Intro
00:44 - What 70 factory visits taught him about deployment vs. demos
02:47 - No SLA in a research paper - why factories are a different game
04:23 - Why he put a packaging machinery veteran in the COO seat
06:34 - The "Useful Now" thesis and where the robotics narrative is wrong
08:53 - The Tesla vs. Waymo parallel for robotics
10:01 - You can't buy your way into a large enough manipulation dataset
10:27 - Why vision alone isn't enough for industrial tasks
12:54 - The pen-in-a-bin problem: why vision-only models are too slow
14:37 - Why robotics is not like LLMs - there is no single scaling law
16:32 - The application-specific full-stack quadrant: why no one else is here
17:12 - Best version of the model-first argument - and how he pushes back
19:50 - What happens to humanoids if "Useful Now" works
21:56 - Inside an electrical prefab shop - what actually happens in there
23:53 - Prefab-Cell-E1: eight robots, $3/hour, 9x productivity
24:44 - What "tailing an outlet" means - the actual task, step by step
28:01 - Wire-bending model generalizing to colors it was never trained on
29:16 - The integration trap: why custom fixtures wreck margins
31:29 - When do you know deployment economics actually work
32:08 - The data flywheel: why 50% success rate is the threshold
33:29 - Touch is filling the gap where vision saturated
35:14 - Combining neural nets with classical control - and why both matter
37:44 - The world action model: image, proprioception, tactile, action, all in
39:39 - You can't buy your way to multimodal data from the internet
40:42 - If this works: data centers on the moon

Factory Intelligence: https://factoryintelligence.com

The OPTIM Update covers real-world AI, automation, robotics, industrial systems and AI Infrastructure for founders, investors, and operators.

Subscribe: https://www.optim.vc

We cover what advanced manufacturing actually is and why it's different from anything most people picture, the sensing gap that's been gating progress for decades, and what cold atmospheric plasma reveals about a material that no other instrument can. Then we get into the commercial reality - who's already deployed, how the business model works, and why the data moat gets harder to close with every passing month.

Jared co-invented the core technology at UC Berkeley alongside his co-founder and CTO Ali Mesbah, a tenured professor who left to build SirenOpt. Their customers include Tier-1 manufacturers across turbines, batteries, and semiconductors in North America, Europe, and Asia.

Learn more about SirenOpt: https://www.sirenopt.com

Volta Foundation presentation mentioned by Jared:
https://youtu.be/a4aW8uxGMQk?si=MGQJI7hNhzzf5pfH

The OPTIM Update covers real-world AI, automation, robotics, and AI Infrastructure for founders, investors, and operators.

Subscribe: https://www.optim.vc

In this episode we cover:

• What it actually feels like to spend a day on a heavy industrial job site
• Why humanoid form factor is the wrong frame for most of these applications
• The self-driving car parallel: we didn't solve autonomous driving by putting a robot in a taxi
• Why "simple" tasks like drawing a line or drilling a hole are deceptively complex
• The tribal knowledge problem - losing the human LLMs built over decades on job sites
• What a job site looks like in 2030
• The shipbuilder who got 2 applicants for "welder" and thousands for "robot welding technician"

Raise Robotics: https://raiserobotics.ai/

The OPTIM Update covers real-world AI, automation, robotics, and AI Infrastructure for founders, investors, and operators.

Subscribe: https://www.optim.vc

We cover how models are actually built - the training pipelines, the data, and what separates people who've done it from people who've read about it. Then we get into deployment reality - what happens when AI hits a real factory floor. And we close with the part nobody wants to discuss - how do you actually sell this stuff through integrators and OEMs.

Asad previously worked at Vicarious (acquired by Google) and ByteDance's robotics program. Trener recently raised a $32M Series A co-led by Engine Ventures and IAG Capital Partners to scale their Acteris platform - a robot-agnostic AI skills platform that works across ABB, Universal Robots, FANUC and others.

Learn more about Trener Robotics: https://trener.ai

The OPTIM Update covers real-world AI, automation, robotics, and AI Infrastructure for founders, investors, and operators.

Subscribe: https://www.optim.vc