Interesting, military only gave ~$1M for such an ambitious project while Uber is burning billions a year and could easily afford this.
My speculation is that this partnership is first and foremost an attempt to provide Uber with a patina of respectability for this attempt and access to military technical expertise. Uber just crashed and burned on self driving, something it desperately needs to quench its towering money fires. Uber's investors know that. Therefore, Uber needs to constantly appear to be "innovating" in flashy ways to keep its investors at bay with seductive promises of even greater riches.
This, even if the entire idea of flying cars driven by normal humans is insane or worse, piloted by a company known for breaking rules (now in Aviation of all things!) and have the dubious honor of being the first to kill a woman with self-driving AI.
It helps to have connections to get these contracts yo. And of course, yet another major American company cooperating with defense. Military-industrial complex indeed. Every peaceful venture must be tied to the killing machine. All the cool kids are, like Google and its TF support for drone assassinations.
This is a cooperative research and development agreement ("CRADA"), which is generally a technology sharing vehicle. Money flows from the non-federal party to the federal party on these agreements. Looks like ARL is matching the funds, so Uber is putting $500k on the CRADA and ARL will reciprocate with $500k in sole-source funding on an actual contracting vehicle. I'd be surprised if the $500k paid for anything more than a proof of concept demonstration on both sides.
I have done multiple CRADAs and frankly believe they are the most underutilized tool in the tech industry, they're the intellectual equivalent of buying mining rights for $5/acre.
The government has staggeringly interesting problems that need solving and sole access to some very interesting data, which they will happily share as long as you bring your own pick-axe and shovel.
Go to federallabs.org and start shopping. It's not Amazon, you'll have to actually go to some of these places and meet the custodians of the problems and data, but it's worth it.
I've said Google is where the future falls from the sky like rain. These labs are like the Mississippi River basin, all the water goes here.
The key is prospecting. You want to look around, get a good idea of how big a particular reserve is, and go for something big. You may have to start small, but keep your eye on the prize and work up a glide slope. You can get there. The law was created so you can get there.
Key programs
* Veterans Affairs
* DoD: too many labs to name, coordinated by each service, ONR, AFRL, and ARL
* NOAA
* DoE
* DoI
* DHS
Really, get a look at the Death and Taxes Poster if you want to shop by category. But the labs are the shelves lined with product you need to shop through.
This sounds super interesting but my knowledge in this area is too limited to fully get what you're saying.
they're the intellectual equivalent of buying mining rights for $5/acre ... The key is prospecting. You want to look around, get a good idea of how big a particular reserve is, and go for something big.
Could you elaborate on this a little or give an example?
As far as I understand CRADA is "cooperative R&D", so each side contributes to the R&D of some subject. Wikipedia says
"Private corporations participating in a CRADA are allowed to file for patent, and they retain patent rights on inventions developed by the CRADA. The government gets a license to the patents"
So if the govt already gets a license to the patents, and the tech has been developed with defense purposes in mind, how am I going to profit as a private company?
So if the govt already gets a license to the patents, and the tech has been developed with defense purposes in mind, how am I going to profit as a private company?
Let's break that down a bit...
So the government has patent rights. They'd have to exercise them, which requires product development. Are you aware of the government selling many widgets to citizens? They sell road maintenance, insurance, and security services. But they don't really sell citizens "things".
Companies are savvy. If the government approached you and said "we have this patented tech, we want you to use it in this widget", now that you know about CRADAs, wouldn't you be a bit worried about a lawsuit from the other people who have rights on that patent?
Further, without commercialization, the government has rights on that one thing. They don't have rights on version 2, or even version 1.1.
You mention defense. Why? Are you in the defense business? Dept of Transportation might be better for self-driving cars. DoE or NIST might be better high tech partners. USDA or FDA if your in the food space. Etc.
Ok, I get the patent part. But what I was mostly interested in is how you see CRADAs help companies make money.
To me CRADAs just sound like a vehicle to share R&D costs, which, in most companies, aren't dominating the budget. So they can be a nice help but no game-changer.
However you mentioned "mining rights" and "going for something big", so it sounded like there's another way to look at this, which I was interested to hear about.
It'll take 6 months to a year to get a CRADA written and signed. They're good for 5 years and can be amended. Budgets are generally side by side, transfer of funds doesn't work well. If you need people on site, working through a DoD staffing agency (e.g. Booz-Allen, one of the military medicine foundations, etc) is worth the overhead because they have the staffing experience.
There are tons of how-tos on line. Keywords are "CRADA", "technology transfer" (T2) and pick a noun or verb of your choice. Every federal department and agency has some sort of "T2" program with people who will even help write the CRADA for you. Commerce, Interior, HHS, all of them.
The key items are to write an effective statement of work and a budget. And understand these things are designed to accomodate big, enormous projects. Act of Congress scale projects. So, if you're going to think, you might as well think big. The paperwork is the same, big or small.
"Interesting, military only gave ~$1M for such an ambitious project"
I imagine this is just a test. They are blowing the cash with 99% certainty nothing will come out of it. BUT, if headway is made, it could quickly become hundreds of millions.
According to Gizmodo report military is using Google's Tensor Flow units for footage processing, it would be fair to assume that it's simply a trial contract to see if they can process some of that footage.
That could potentially mean government contracts for private companies going to the lowest bidder, some time in the near future.
And of course, yet another major American company cooperating with defense.
Just wanted to point out that a large chunk of technological innovation in the past century has roots in the military. Almost all communications protocols had origins in defense.
Not really - a $1 million CRADA doesn't make a company's existence important or guaranteed in the eyes of the military. If the technology is deployed, different story.
Comments
Interesting, military only gave ~$1M for such an ambitious project while Uber is burning billions a year and could easily afford this.
My speculation is that this partnership is first and foremost an attempt to provide Uber with a patina of respectability for this attempt and access to military technical expertise. Uber just crashed and burned on self driving, something it desperately needs to quench its towering money fires. Uber's investors know that. Therefore, Uber needs to constantly appear to be "innovating" in flashy ways to keep its investors at bay with seductive promises of even greater riches.
This, even if the entire idea of flying cars driven by normal humans is insane or worse, piloted by a company known for breaking rules (now in Aviation of all things!) and have the dubious honor of being the first to kill a woman with self-driving AI.
It helps to have connections to get these contracts yo. And of course, yet another major American company cooperating with defense. Military-industrial complex indeed. Every peaceful venture must be tied to the killing machine. All the cool kids are, like Google and its TF support for drone assassinations.
EDIT: added technical expertise to 2nd paragraph
This is a cooperative research and development agreement ("CRADA"), which is generally a technology sharing vehicle. Money flows from the non-federal party to the federal party on these agreements. Looks like ARL is matching the funds, so Uber is putting $500k on the CRADA and ARL will reciprocate with $500k in sole-source funding on an actual contracting vehicle. I'd be surprised if the $500k paid for anything more than a proof of concept demonstration on both sides.
I have done multiple CRADAs and frankly believe they are the most underutilized tool in the tech industry, they're the intellectual equivalent of buying mining rights for $5/acre.
The government has staggeringly interesting problems that need solving and sole access to some very interesting data, which they will happily share as long as you bring your own pick-axe and shovel.
Go to federallabs.org and start shopping. It's not Amazon, you'll have to actually go to some of these places and meet the custodians of the problems and data, but it's worth it.
I've said Google is where the future falls from the sky like rain. These labs are like the Mississippi River basin, all the water goes here.
The key is prospecting. You want to look around, get a good idea of how big a particular reserve is, and go for something big. You may have to start small, but keep your eye on the prize and work up a glide slope. You can get there. The law was created so you can get there.
Key programs
* Veterans Affairs
* DoD: too many labs to name, coordinated by each service, ONR, AFRL, and ARL
* NOAA
* DoE
* DoI
* DHS
Really, get a look at the Death and Taxes Poster if you want to shop by category. But the labs are the shelves lined with product you need to shop through.
https://www.timeplots.com/products/death-and-taxes-2016
This sounds super interesting but my knowledge in this area is too limited to fully get what you're saying.
they're the intellectual equivalent of buying mining rights for $5/acre ... The key is prospecting. You want to look around, get a good idea of how big a particular reserve is, and go for something big.
Could you elaborate on this a little or give an example?
As far as I understand CRADA is "cooperative R&D", so each side contributes to the R&D of some subject. Wikipedia says
"Private corporations participating in a CRADA are allowed to file for patent, and they retain patent rights on inventions developed by the CRADA. The government gets a license to the patents"
So if the govt already gets a license to the patents, and the tech has been developed with defense purposes in mind, how am I going to profit as a private company?
Let's break that down a bit...
So the government has patent rights. They'd have to exercise them, which requires product development. Are you aware of the government selling many widgets to citizens? They sell road maintenance, insurance, and security services. But they don't really sell citizens "things".
Companies are savvy. If the government approached you and said "we have this patented tech, we want you to use it in this widget", now that you know about CRADAs, wouldn't you be a bit worried about a lawsuit from the other people who have rights on that patent?
Further, without commercialization, the government has rights on that one thing. They don't have rights on version 2, or even version 1.1.
You mention defense. Why? Are you in the defense business? Dept of Transportation might be better for self-driving cars. DoE or NIST might be better high tech partners. USDA or FDA if your in the food space. Etc.
Ok, I get the patent part. But what I was mostly interested in is how you see CRADAs help companies make money.
To me CRADAs just sound like a vehicle to share R&D costs, which, in most companies, aren't dominating the budget. So they can be a nice help but no game-changer.
However you mentioned "mining rights" and "going for something big", so it sounded like there's another way to look at this, which I was interested to hear about.
That's super interesting. Would you be willing to share some of your background on this? I'm quite curious about this route.
It'll take 6 months to a year to get a CRADA written and signed. They're good for 5 years and can be amended. Budgets are generally side by side, transfer of funds doesn't work well. If you need people on site, working through a DoD staffing agency (e.g. Booz-Allen, one of the military medicine foundations, etc) is worth the overhead because they have the staffing experience.
There are tons of how-tos on line. Keywords are "CRADA", "technology transfer" (T2) and pick a noun or verb of your choice. Every federal department and agency has some sort of "T2" program with people who will even help write the CRADA for you. Commerce, Interior, HHS, all of them.
The key items are to write an effective statement of work and a budget. And understand these things are designed to accomodate big, enormous projects. Act of Congress scale projects. So, if you're going to think, you might as well think big. The paperwork is the same, big or small.
Great advice. Thank you.
"Interesting, military only gave ~$1M for such an ambitious project"
I imagine this is just a test. They are blowing the cash with 99% certainty nothing will come out of it. BUT, if headway is made, it could quickly become hundreds of millions.
According to Gizmodo report military is using Google's Tensor Flow units for footage processing, it would be fair to assume that it's simply a trial contract to see if they can process some of that footage. That could potentially mean government contracts for private companies going to the lowest bidder, some time in the near future.
tensorflow is a machine learning API. you're thinking of TPUs
Just wanted to point out that a large chunk of technological innovation in the past century has roots in the military. Almost all communications protocols had origins in defense.
Well maybe its a cheap way of getting access to the research larger drones for resupply in the field
I guess they’re starting their IPO roadshow by trying to stay in the news as much as possible.
With as flimsy as this is, is it an ICO* roadshow?
That, and they just made it militarily important that they continue to exist.
Not really - a $1 million CRADA doesn't make a company's existence important or guaranteed in the eyes of the military. If the technology is deployed, different story.