A robot arm consumes 21,950 kWh annually, on average. If we wanted just 1M more robots making things in the USA, we would need nearly 22B kWh just for the robots.
Now imagine we wanted to reindustrialize more.
Our grid is barely keeping up today.
Nuclear is the only answer.
The US installed 39,576 robots in 2022.
The US has at most 39,169 robotics engineers working in manufacturing.
Coincidence?
This isn't a call to just say we need more people studying robotics (we know we do).
Robots and automation hardware needs to be easier to integrate and
@APompliano
History is rife with people mocking the value of new platforms. Whenever I'm tempted to do so, I recall quotes like this:
"Well-informed people know it is impossible to transmit the voice over wires. Even if it were, it would be of no practical value."
- Boston Post, 1865
@APompliano
Or this all time favorite:
"This 'telephone' has too many shortcomings to be seriously considered as a means of communication. The device is inherently of no value to us."
- Western Union internal memo, 1877
Even before the US Steel sale, the USA only produced ~80M tons of steel in 2022, or 4.2% of global production. By comparison, China produced 1.017B tons of steel, or 53.9% of global production.
How much steel would we need to be producing to build a proper O'Neill cylinder?
@Scobleizer
Batteries need to get a lot better IMHO. Solar definitely getting pretty cost effective. Nuclear generates power 24/7 though and we know it works. It'll take years for new reactors to come online, which is why I think we should be building now.
We're entering the era of general purpose robotics.
Robots are moving beyond the "mainframe" era of big, fixed installations.
Companies are starting to deploy robots for applications outside of their traditional use in high-throughput manufacturing.
Here are a few examples:
Tooling and parts presentation literally make or break manufacturing operations, especially if you ever want to do anything with robots. Also true when humans are in the loop, but especially true with robots.
1 product could have 50,000 parts, each part may have 10-20 operations to convert it from raw material to finished
Each operation requires a jig or tooling for repeatability
Here's another great example of the type of flexible robotic automation powered by our technology at
@ReadyRobotics
: one of our TaskMate systems with a
@Universal_Robot
arm tending a
@FlexArmInc
tapping arm. Setup only took hours, instead of days or weeks!
The team at
@drivecapital
are the definition of contrarians: they were investing and building in Columbus before anyone thought it was hot.
Also - if you're an industrial startup, the Mid-West is where you want to be, and you can't get much more central than Columbus
In 1979, about 580,000 computers were sold. Just over ten years later, more than 21,000,000 computers were sold.
In 2019, about 373,000 robot arms were sold. I won't be surprised if a similar 36X increase is achieved over the next decade.
@paulg
@KTmBoyle
It's even stranger when you stop and think about the fact that so many truly groundbreaking lines of scientific inquiry happened outside the normal academic tracks. Einstein. Goddard. The Wright Brothers.
I'm biased (since I'm at one of their portfolio companies), but
@EniacVC
invests in companies working on awesome tech in interesting spaces. So check 'em out!
DEAR ROCKSTARS LOOKING FOR YOUR NEXT BIG CAREER MOVE:
Join the
@EniacVC
Talent Network! 👩🏾💻👩🏻🎓
Members get direct access to Eniac’s portfolio job openings and private intros to top companies.
Request access here:
ForgeOS is the bridge between bits and atoms.
Case in point: our modular robotic palletizing system.
ForgeOS controls the hardware, and the HMI running the READY Pallets app controls ForgeOS via REST APIs. We launched this in July and just closed the contract for the first
Get in, we're exascaling the American industrial base.
@atomic_inc
closes $17M seed to reproduce human manufacturing skill via AI. The fundraising will continue until acceleration improves!
@abemurray
Thanks
@abemurray
! Graphic is my own from gen AI plus a little old-fashioned Photoshop polish. My personal opinion is the idea funnel is actually bottlenecked by complexity in the following way...
THIS––is the FUTURE of all SCIENCE.
remote-controlled robots...🤖🦾
automatically doing experiments...🧪⚗️
designed by scientists on laptops...👨🏽💻 🧑🏾🔬 👩🏽💻 👩🏾🔬
run on a creative cloud––from anywhere... 💭 🌍
👇👇GENIUSES at
@StrateosLab
Exciting News! After months in stealth, we're introducing Dirac's BuildOS - the first AUTOMATED Work Instruction platform.
It's time to change the way manufacturing engineers draft work instructions.
Get access to BuildOS here:
1/9
"[Since 1974], 42% of public companies are venture backed, representing 63% of total market capitalization. These companies account for 35% of total employment and 85% of total R&D.
That's pretty good for an industry that invests about 0.4% of GDP."
-
@skupor
🤯
@codyaims
Tooling and parts presentation literally make or break manufacturing operations, especially if you ever want to do anything with robots. Also true when humans are in the loop, but especially true with robots.
I'm really excited to share more details soon about the next version of Forge/OS launching this spring. It addresses the biggest unsolved problem in robotics today: fragmentation. Every OEM uses their own proprietary programming language. We're fixing that.
@Scobleizer
I think at the end of the day, the right answer is we should be building as much solar and nuclear as we possibly can. We'll never regret having more power.
Thanks
@Forbes
for recognizing us as a start-up to watch in the IoT space! Networking robotic systems together is an important aspect of building integrated smart factories, and just one of the capabilities available in our Forge/OS software platform.
He's right. No real VCs in the medtech or biotech space touched Theranos. They're some of the most savvy scientists out there and no way could Theranos pass muster with them.
Alternate narrative: Silicon Valley VC is actually a defense against this bullshit. Theranos and Fyre never would have gotten off the ground without dumb money from New York and Washington, and fauning coverage from New York/Washington media. Who will fund my documentary?
Many don't grasp how much of a startup
@EniacVC
truly is. For 4 years, we hustled during nights and weekends without drawing a single dollar of salary. I think this journey gives us a very different perspective working with founders than most VCs who join an existing organization
Congratulations to Accufacture for launching their Forge/OS-powered AccuTask 700 and AccuTask 1500. It’s so exciting to see others building around Forge/OS to create powerful robotic solutions that can be programmed by anyone on the factory floor. No training required.
We need a name for ChatGPT style prose, because I swear I can identify it in the wild. It has a certain vibe. Like it was written by a golden retriever.
This is creative destruction, a.k.a. Schumpeter's Gale, from Austrian-American economist Joseph Schumpeter: the "process of industrial mutation that incessantly revolutionizes the economic structure from within, incessantly destroying the old one, incessantly creating a new one."
Extremely proud to announce
@a16z
’s investment in Castelion to shore up American hypersonic capabilities and bring peace through deterrence. 🇺🇸
This team moves faster than we ever thought possible. If you’re looking for an important defense mission where every single day
Don’t let them fool you.
We don’t need cheaper labor. We need smarter labor.
Humanity evolves under time and labor constraints.
We will defeat and surpass those powers who so badly want us to fail. 🇺🇸
This article perfectly encapsulates the state of the robotics industry today. We're entering the era of general purpose robotics, which is why we're working on Forge/OS: an enterprise-grade OS that enables anyone to program any robot arm with the same interface.
I see this truth every day in the plants that deploy automation. More automation leads to more people being hired as work increases at the plant.
The best plants don't even fire anyone to begin with because they know how hard it is to find competent people in the first place.
@Noahpinion
@cstross
actually created one of the atypical dystopias in Accelerando with predatory machine intelligences ruthlessly focused on debt collection (amongst other things)
This is the way. Machines building machines has already been the unlock of the last three industrial revolutions that gave some of humanity an embarrassment of riches.
We just have to keep doing it and go faster and faster and faster.
We have to build the machines that build for us.
We need the operational and intellectual leverage of robotics & AI.
It's the only way to really accelerate our scaling and efficiency.
This report raises many interesting points about the wave of automation sweeping through the economy, including the yawning skills gap being faced by many industries. The manufacturing industry has long felt the cost of this skills gap.
This has always been one of my favorite visualizations of our solar system. I wish there was one that zoomed out further to show the sun orbiting the supermassive black hole at our galactic center.
@kscottz
I don't agree that it's nationalism to want to strengthen the US manufacturing base, or to point out that certain critical items should be manufactured domestically, especially in the case of items used in national defense (like computers).
The sleeping giant must awaken, with renewed purpose and unshakeable resolve.
Per this Politico article, we need to be manufacturing a LOT more domestically for defense:
Once you create standardization at one level of your technology stack, those standard components become the building blocks of the next wave of innovations.
This is why we've built standardized interfaces for robots and industrial automation with ForgeOS at
@ReadyRobotics
. It
Imagine a world where there are hundreds of millions of robot arms sold annually instead of hundreds of thousands.
Computers helped us automate the organization of electrons at scale. Robots will help us automate the organization of atoms at scale.
Numbers for the robot installations comes from the IFR's World Robotics Industrial Robots 2023 report.
Stats on the number of robot engineers comes from the US Department of Labor's O'Net Online database:
Like all statistics, these job numbers are
@sethbannon
In fact, when I used to do university tech transfer, we'd hunt for exactly those sort of people to become the technical foundation of new companies. And some universities include patent/licensing activity as a measure for tenure decisions now, not just publishing.
To quote
@FutureJurvetson
: "robots, drones, satellites and self-driving cars are the kinds of things that excite me." Which is why I'm excited to talk with
@rayfega
about all things robotics tomorrow at 5pm CT / 8pm ET. Come join us!
A new customer had six aluminum plates, each with 600 holes that needed to be tapped. Other shops said it would take months. See how Little Enterprises was able to do it in a week.
Germany doesn't have this problem because their society doesn't look down on manufacturing jobs. We need to do better in the US. Sone of the smartest and most talented people I know work in factories.
There are many pathways to a rewarding career. With 2.4 million manufacturing positions expected to remain vacant in the next decade, it's time to recognize that not all careers require a traditional four year degree.
Windows ignited the PC Age. iOS unlocked the Mobile Age. Tomorrow, Forge/OS 5 unleashes the Automation Age. Join the Livestream Launch Event and experience the future of automation:
It's not tool use that sets humans apart from other life. Chimps use tools. So do crows. Nor is it the building of structures. Bees build structures. So do termites. Nor domesticating other life. Ants herd aphids.
It's the building of machines that ultimately sets humans apart.
Here's another great example of a robot powered by our software, FORGE/OS. When you use FORGE/OS, setting up a robot is so intuitive that factories capture ROI with small production runs of a few hundred parts, instead of millions.
I think part of the reason so many US entrepreneurs are building companies where the product is software or new business models is that that's the stuff our culture seems to value. But we could *really* use more plays in hardware and industrial process improvement
In 1830, a farmer could harvest 0.2 hectares of grain a day using a handheld scythe. In 1831, the same farmer could cut 2 hectares a day, a 10X increase. How?
The McCormick reaper. Physical automation led to an explosion in productivity.
@KennethCassel
Arguably none of them? I am biased though since we make software that enables you to program industrial robots in a uniform, easy-to-use way.
If industrial robots were easy to use though, my company wouldn't need to exist.
I couldn't agree with this sentiment more. Automation tends to create more jobs in the factory than they replace and few factories actually let workers go. Instead they get re-assigned elsewhere in the facility.
NY Times pushing robot tax based on the assertion that automation is designed “merely to take away jobs.” The real threat to jobs is the inability to compete. Robots help make US companies competitive in every industry, contributing to our strong economy and low unemployment.
The future is already here, if you know where to look. 24-hour timelapse of the world's largest autonomous drone delivery system, covering an entire country.
I would also add to
@kevg1412
's list: the Cohen-Boyer discovery of recombinant DNA that was patented and licensed out by Stanford in a very intentional way to foster growth of biotech industry...the very first licensee was Boyer himself for his little startup called Genentech!
Very surprised by Naval's take here. Studying Silicon Valley history it's clear that tech is there for three major reasons (among others):
1) William Shockley moved to Palo Alto to be close to his sick mother. There, he set up Shockley Labs and recruited tremendous talent to
The first robot programming language was VAL: Variable Assembly Language, used to program the Unimate robots. Since VAL's launch in 1961, robot programming languages have splintered, with dozens today. This fragmentation holds our industry back.
@abemurray
But I won't pretend that what we're building is a magic panacea for the whole industry, it's just one part of the toolchain. There are many other things that need to be improved, to name just a few more:
1. Simulation needs to be made much, much better.
2. Grasping for robots
@KTmBoyle
@terronk
@micsolana
Groundhog Day. Which now that I stop and think about it, also explains why I loved Edge of Tomorrow so much when I first saw it.
@abemurray
Imagine you're a software developer and you want to get into robotics. You're used to a world where you can learn some APIs and push code that just runs on almost every PC or cloud instance out there. Then you look into robotics and what do you find? Endless, stupid complexity.