This article is a quick memo to address questions from the audiences:
- Learners: How do I know it’s the right robot for me?
- Researchers: Why should I choose an open robot over an expensive one?
- Tech Partners: How do I know SCUTTLE designs will be supported in 5 years?
- Tech integrators: How do we know it’s truly research-grade tech?
- Software Developers: Why get involved with SCUTTLE since it’s capabilities don’t seem dazzling?
SCUTTLE core designs will always have the best technology.
- Nobody makes better batteries than Panasonic and Samsung. That’s why we leverage their cells, and we keep testing forever.
- Nobody made a more popular communication interface than the 40-pin header popularized by Raspberry Pi. That’s why we designed our wiring & communications around it.
- Nobody offers a better OS than Linux. And if they do, it will be opensource just the same. So we use it.
- Nobody made an easier roadmap to AI than Viam. That’s why we partnered with their engineers. Recall MS DOS vs Windows? That is like ROS vs Viam and the new one is easier.
- Nobody makes a frame that’s stronger than ours, which can still be adapted in size without a cost impact. That’s why we designed it the way it is.
- Nobody unlocks AI at a better performance/cost ratio than Texas Instruments. That’s why we integrated TDA4VM Edge AI together.
- No investor owns our designs. They belong to the world. It can never be paywalled. Our software can’t go out of date because it rests atop open scripts by companies with 50+ years of growth.
- Nobody has longer lifecycles across their robot hardware than our tier-twos. Ninety percent of our hardware stands in production over 30 years. These elements are all in mass production, backed by tier-2 manufacturers ready to support every customer & engineering team.
- Nobody open-sources full-spectrum open-source tech than SCUTTLE. Examine the wheel assembly, encoders, wiring, or power system of a turtlebot (most popular research platform as of 2023) – it’s hard; they only opensource the software.
- Nobody loses when we share designs, including us. Imagine if Dominoes pizza open-sourced their pizza recipes – how much market share would they lose? Not much. Customers want to pay for the pizza, not for the recipe.
And, to address the last question: between 2017 and now, we have focused on robustness of the platform, and making things modular enough to always have the best capabilities, even if the best changes each year.
For dazzling capabilities, you might desire a robot with an arm (aka mobile manipulator) so we continuously scan the best arms each year, and making sure our platform supports all of them; having sufficient power, communication lines, payload, and dynamic ability to stay balanced with a moving end-effector. The design is not static. It’s intended to customize easily whether you add a $200 arm or a $20,000 arm.
…but does Pansonic really work for us?
This section added 2023.11.13
In a big way, these companies (panasonic, TI, etc) do work for the SCUTTLE project. They create value that we use directly. The more aligned our interests, the more these companies outputs serve our designs with precision.
What is better than an engineer working for you? An engineer working in a company that works for you. The company offers pay, benefits, all overhead costs, and a supporting team as well as training so the engineer can crank out value. Then, we must ask about direction.
What good is it to have these engineers working for someone else’s dream? Well, “Great minds think alike” means that “Great dreamers dream in unison.”
We know the flavor of improvement in robotics and the designers of our components know the same. An example of this leveraged effort from outside of a company is Denso and Toyota. Denso is so advanced in making components, Toyota wouldn’t dream of developing their own competing devices of the same type, such as fuel injectors – extremely precise electromechanical components at the core of the engine, which is at the core of the vehicle, which is at the core of Toyota’s business.
(just to drive the point, if your fuel injectors release the fuel for 1/1000 of a second too long, the car can go from passing emissions to failing emissions) and this injector must operate under extreme pressure and several hundred degrees of temperature plus vibrations for 10 years – that’s literally the USA’s legal requirement for EPA emissions warranty.
The great size of Denso gives it the capacity to have its own research teams, funding crews of PhD brains to create the next generation of parts in the component. Instead of paying for those teams (research is a cost, not a revenue), Toyota enjoys the benefits of the technology to make the very best engines. Suppliers like this are so advanced that they are developing technology that informs lawmakers, such that when the US government passed a law to require 10% ethanol fuel (made from USA corn crops) it very well could be because the folks at Denso made a recent breakthrough that allowed the fuel system to survive 10% corn, and maintain performance. (Yes, your gasoline is 10% corn and we calibration engineers still made it run!) In the meantime, Denso is also selling their components to Chrysler, Ford, General motors, and more automakers. Those sales drive Denso’s revenues to continue innovating and stay at the head of the pack.
So let’s return to the situation of Toyota’s interests – how can I benefit from the wonderful engineering of my supplier who also sells their technology to my competitors? Well, this is where ownership can MAYBE come in. Toyota owns around 25% of Denso, giving them some sway in the patterns of distribution. The tradeoff is simple: we pay a lot if we want exclusivity, but if everyone gets the technology then the research pays for itself – that’s how the best remains the best. Exclusivity is the only thing that must be paid for, and it’s a dangerous road.
The path of exclusivity is the path of Blu-ray and HD-DVD. You might want your technology to be proprietary (as does your competitor) but two massive forces act against you:
1) your own desire to have your tech adopted globally, which is fairly opposite to the definition of exclusive, and
2) the fact that the entire world has an interest in paying the best price, which has no extra margin for your brand exclusivity. Your small company leaves the rest of the 7 billion people to figure out a way to geet the same value but not pay a margin for your precious brand.
So, in the short term (5 years-ish) you perform a sprint, spending loads of money to catch the edge of the tech wave, endangering your workers with pressure to take shortcuts, and you risk losing 100% of that funded effort if you don’t make the very best value option. If you read this AI generated summary, you’ll find that Sony won the game because they had stronger volume in the short term. The market-talkie people call this “wide adoption” but from Sony’s point of view, they jammed a blu-ray into the playstation where the customers were already adopting in high volume – nobody “adopted” the Playstation 3 because they liked the blu-ray technology. PS3’s value would have been the same with the HD-DVD technology inside.
In summary, the blu-ray/HD-DVD story was one where each player was guaranteed a full win or full loss because each player was greedy enough to demand exclusivity. Greed lives at the top, not at the bottom. The engineers at either company would have LOVED to opensource their technology if they knew it would guarantee adoption and their ten years of work wouldn’t go in the trash.
On the other hand, operations like Linux cannot ever be competed with, as the technology leverages the entire world’s contributors. And these contributors know it creates so much value that their benefits, just in functionality, make it worthwhile to build improvements without getting a paycheck for their contribution.
Think about all of the brilliant designers who have reasons to contribute:
- They’re in between jobs and they want to keep their skills honed
- Their skillset is in extreme demand, but their citizenship doesn’t grant them work permissions in the country with the employer.
- They performed amazing development with in a bleeding-edge company, and they got burned out and took family leave but missed some coding time.
- They saw their brilliant technology getting used inefficiently by a company’s leadership, and preferred to build tech on their own.
- They had an expensive skillset that could only be afforded by departments of defense funding and they didn’t morally align.
- They met a wonderful spouse who earned enough money, relocated, and still wanted to put their skills to use in a remote place where their field is not hiring.
- Their designs have more broad application than the niche where their company is leveraging it, and they want to keep their paycheck while seeing their work create more benefit to humanity.
- They know how to pivot into a new type of tech but their employer doesn’t, and they had better get a head start before their current industry fizzles.
I could name dozens more situations where extremely smart and educated people would want to contribute in open source projects, including those minds behind the most expensive and exclusive technology like Apple and Google products. They are humans with their own dreams and the employers who operate on the intellectual property can have plenty of property but they can’t own the intellectual.
If we build a structure (a set of robot modules, and an open source community) that can employ all of the passion and designs of great people, then we are not required to contractually confine them to our business, and we can still use their work to build advanced designs for the next decade. This is the essence of open source inventions.