One of the problems the Navy has run into in recent years is
intellectual property rights. Understandably, manufacturers don't want to give
up their proprietary data. Equally understandable, the Navy wants exactly that
data so that they can model system performance as they contemplate system integration
and modifications.
It's an impasse, currently, and has resulted in the Navy abandoning some
promising systems.
The seriousness of the problem is magnified by the Navy’s [unwise] attempts to substitute modeling for actual tests in order to save money. DOT&E has documented this problem multiple times across multiple weapon system programs in their various annual reports.
Consider some of the implications of this issue. What about the manufacturer who was asked to provide proprietary data for the 32 Zumwalts. Hmm … a contract for 32 ships might be worth giving up the data. Unfortunately, the production run was reduced from 32 to 3. Is it still worth it to the manufacturer to potentially give up their hard won institutional knowledge that gives them a competitive advantage for just 3 ships?
The Navy is erratic, to say the least. Those 20 FFG may become 40 or the Navy
may decide to go whole hog on unmanned and cut the FFG at 8. Who knows? The
point is that the manufacturer is being asked to give away their proprietary
information for the hope of a substantial sale that history strongly suggests
is unlikely to occur.
Further, the Navy's ability to protect proprietary data is highly suspect with
network infiltrations happening all the time in addition to whatever unethical
and improper actions the Navy might opt to take that gives away the
manufacturer’s data.
If I were a manufacturer, I'd be very reluctant to give away my hard-earned industrial
advantage for a shaky deal with the US Navy.
Some companies might consider it worthwhile but, apparently, many are
deciding it's not worth it.
Consider the following specific example from a DOT&E report.
In FY15, DOT&E learned that the Program Executive Office for Integrated Warfare Systems (PEO IWS) stopped work on the PRA [ed. Probability of Raid Annihilation] Test Bed for the Freedom variant because a high-fidelity model of the ship’s AN/SPS-75 radar was not being developed. Development of an acceptable radar model requires intellectual property rights that the Navy does not hold … the Navy has also been unable to develop a high-fidelity model of that ship’s AN/SPS-77 radar for the same reason.[1, p.226]
Instead of conducting realistic, live tests, the Navy wants to model and simulate self-defense capabilities for the LCS. That’s a very poor approach. Modeling and simulation are fine but as an adjunct to live fire testing, not as a replacement for it. That aside, in order to construct a valid model the Navy needs all the data on the radars. The radar manufacturers, however, refuse to provide the proprietary data. This leaves the Navy unable to program realistic models and simulations.
It is easy to see, then, that the Navy’s choice of future equipment (of all types) is limited to those for which they can obtain proprietary data. This eliminates many fine pieces of equipment from manufacturers who are unwilling to expose their data.
________________________________
[1]Director, Operational Test & Evaluation, “FY 2015 Annual Report”, Jan 2016
I'd divide things into two categories. If the radar (for example) was already developed by the manufacturer, at their expense, I can certainly sympathize with them over not wanting to give up the technical data. But the Navy actually pays for the development of a lot of stuff. I suspect, though I'm not certain, that the Navy (or maybe Defense Dept. as a whole) paid for the development of the Aegis radar, didn't they? I'm pretty sure Raytheon didn't develop the SPY-6 radar at their own expense and then hope to sell it to the Navy. In that case, isn't it a little more reasonable for the Navy to demand rights to the technical data?
ReplyDeleteYou're quite right. If the Navy pays for the development of a product specifically for their use then, of course, they should own the data rights, if they wish. However, THEY MUST SPECIFY THAT CONDITION IN THE PURCHASE ORDER.
DeleteMost of what this post refers to involves much smaller products that were, actually, developed by the manufacturer. For example, Thales CAPTAS variable depth sonars, TRS-3D/4D radars, various EO/IR sensors, and so on. Those are just products on the market and the Navy had little or nothing to do with developing them and, therefore, no expectation of obtaining performance and design data. Of course, the solution to that is for the Navy to perform actual, live testing and develop their own data sets ... which they seem loathe to do.
Alternatively, the Navy could offer to buy the data from the manufacturer but, if I were the manufacturer, knowing the Navy would use the data to bypass me, cut me out of purchases, and release the data (intentionally or inadvertently) to other manufacturers and that I'd lose a significant market advantage, I'd charge the Navy an arm a leg, and most of the national debt to do so!
There are also gray areas. For example, CIWS was already a developed product (I have no idea who owns the data rights) and the Navy requested an add-on missile (SeaRAM) or add-on optics. That wouldn't entitle the Navy to the entire data set for the base CIWS just because the Navy funded an add-on. This is just an example. I don't actually know who paid for CIWS add-ons and development or what the data arrangements are/were. I'm just illustrating the concept.
The Navy could always make it a condition of the contract that the IPR was supplied to the Navy, would have thought in most cases they would find a contractor willing to accept those conditions, though expect would be rare exception.
DeleteRemember Fincantieri refused to give the Canadians the IPR of the FREMM for the CSC frigate contract, the Canadians disallowed their bid.
"make it a condition of the contract"
DeleteThat makes for a nice comment statement but it ignores reality. If you were a manufacturer selling your product worldwide, would give away your current and future development and marketing advantages just to satisfy a US Navy that, far more often than not, winds up hugely downgrading the purchase (32 Zumwalts to 3, for example, or 55 LCS to thirty something, or the LCS ASW module that was then cancelled)? I wouldn't.
"rare exception"
Far from a rare exception, DOT&E annual reports cites it as a common occurrent with specific examples of radars, sensors, missiles, sonars, etc. Industry does not want to give up their hard-earned advantages. In fact, the exception appears to be a manufacturer that DOES want to give up their data.
I'm more optimistic and you underestimate what a big stick Navy can wield in the $billions it spends, it needs to use it, will not happen overnight but take years of continuous unrelenting pressure and sure their will be exceptions.
DeleteThe 2020 policy of the Joint Requirements Oversight Council (JROC) pushing hard for open-systems software to ensure component interoperability in its new systems and minimize non-proprietary software eg Boeing have said the E-7 is not an open-architecture system but could be modified to have one if USAF buys the AEW&C aircraft.
"I'm more optimistic and you underestimate what a big stick Navy can wield in the $billions it spends"
DeleteThis is not an issue of optimism or pessimism. It's purely about reality. The documented (see every DOT&E annual report) fact is that the Navy is having a hard time obtaining data and IP rights and has abandoned several attractive technologies and products because of that. This despite throwing billions of dollars at the acquisition programs and contracts.
If a manufacturer simply does not want to give up their data and industrial advantages, no amount of 'continuous unrelenting pressure' is going to change that especially when the Navy repeatedly demonstrates that their promises of mega programs and mega dollars are subject to drastic reductions seemingly on a whim.
Chinese Gordon
ReplyDeleteIsn't it the general opinion here that manufacturers data
Deleteis published in "The Journal Irreproducible Results" ?
As Reagan said "Trust the spec, but still test"
Testing can be expensive, but still necessary. Sadly, the US military keeps falling to "penny wise, pound foolish" behavior, and refusing to perform necessary tests in order to save money- that'll then be wasted fixing problems that would've been identified in testing.
DeleteJust occasionally, manufacturer's data about their products reveals they aren't quite as quite as good as their salesmen claim. This naturally makes them reluctant to provide that data to customers. The problems of the Boeing 737Max are an example. The USN's reluctance to do their own testing looks even worse from that viewpoint.
ReplyDeleteIP is a fun topic. Some is needed to maintain any ship. Construction data for the Zumwalt design? Absolutely required to repair the ship after it suffers damage from the sea or the enemy. Design and construction data for the LM2500? It should be available IF the manufacturer goes out of business. IP for the commercial VHF radio? Hell no, just chuck it overboard (remembering to fill out the environmental impact statement) and buy a new one at West Marine.
ReplyDeleteAs far as buying the IP for modelling purposes, well radar maybe able to tell you the angle of attack of a flea at 110 miles, but how do you model the fact that the OOD and the CIC watch officer aren't talking?
I agree with your point; put it on a ship, take it to sea in good and bad conditions and run some tests. Probably cheaper and more accurate in the long run.
RA
A more apt example of IP might be something like a radar. While the physical radar equipment is important, the underlying software is what determines the radar functionality. The radar receives a return signal as a mash of electronic noise and it is the software treatment of that mash of noise that determines how effective the radar is. The algorithms that decide what blip of noise is a target and what is an anomaly is what's important. For the Navy, that means acquiring the source code which no manufacturer is going to want to give up. Unfortunately, without it, the Navy can't tweak the radar performance, model the radar performance, make predictions about the radar performance against new enemy missiles, ships, and aircraft, can't model electromagnetic interference from self and enemy emissions, and so on.
DeleteA fair comment re the software, but, if I understand your example, an advanced radar would not be sold on the open market but only through a government agency, and probably only with the navy's okay. In which case it is in both the navy's and the manufacturer's best interest to protect the code.
DeleteI would suggest the best way of dealing with this situation is an in-service support contract where the company updates the software as required by the navy, or provide the code if it is not willing to do so. This is actually a win win situation as the navy has a dedicated team modelling and updating the code, and the company maintains the expertise to create the next incremental improvement in the technology. This can keep equipment effective longer; rather than constantly relying on the "next big leap". (which I understand a certain author on this blog is not a fan of.)
RA
"an advanced radar would not be sold on the open market"
DeleteThey're ALL sold on the open market. For example, the LCS TRS-3D/4D and the Seagiraffe are widely sold throughout the world. Even Aegis radar/software is sold around the world and has been bought by Japan, Spanish Navy, Royal Norwegian Navy, Republic of Korea Navy, and Royal Australian Navy.
Agree completely with CNO that we should be doing a lot more actual testing, and not just relying on software simulations. Simulations are fine for initial checks, but they ought to be double checked with measurements.
ReplyDeleteDuring my career (i'm retired now) I was a compute programmer, so I understand more than most folks how error prone the software development process is. In fact, I probably spent more time chasing bugs in existing programs than I did developing new ones. We even have an acronym: GIGO. Stands for "Garbage In, Garbage Out". Most of you have probably heard about that, but may not realize all the types of "Garbage In" that can cause problems. Of course there are errors in the design of the software (the simulation algorithms, for example) and errors in the implementation (coding bugs). But even if the design and implementation are completely perfect (not achievable now and may never be), you can still get garbage out by putting in bad input data. . Which is why simulation is nice, but we should always cross check with actual measurements.
Regrettably, most Americans in general, and the military in particular, place more trust than they should in something that comes out of a computer.
As an aside, this is why I believe that it is delusional to think, as we do, that we can be totally confident about the safety and reliability of our nuclear weapons, and even design new ones, based entirely on software simulations.
DeleteNow it's not quite fair to say that, since we do (I think !) test the individual non-nuclear components and the non-nuclear subsystems. And we might even verify that a metal pit made of a heavy but non-fissile metal will implode correctly under the explosives, using high speed cameras. But we can't verify whether the fission primary will actually "cook off", and we can't verify the coupling between the fission primary and the fusion secondary, without actually doing a nuclear detonation.
My guess? The weapons will PROBABLY work fine. But there is a small but non-zero probability that they won't. Since the stakes with nuclear weapons are so high, this means that there is a small but non-zero chance that we will, at some point in the future, get an awful surprise at the worst possible time.
One point of full disclosure, for anyone who wants to be more optimistic about software development. My experience is in commercial software development (and not even the most careful commercial software development like Microsoft Windows) and not in military software development, which is more rigorous. But that doesn't mean the probability of failure is zero (as we've experienced with all the F35's software problems). It's just smaller than in my experience.
DeleteThis post just highlights how incompetent the Navy Acquistion Workforce has become. They do not think of the total lifecycle of a product, they are only interested in getting a product under contract. BECAUSE of 3 reasons; 1) the long term PMO personell are incompetent and don't know how to ask for and make the case for data, 2) the PM will be gone by time they need to do something that might require data, and 3) if the Program doesn't have the data or rights then they can't be held accountable for not being able to test, model or even support the product.
ReplyDeleteThere are a range of rights that can be granted at contract award: unlimited, Government Purpose Rights, etc. on all or limited portions of the product. But as CNO points out you have to be competent enough to ask for them up front BASED on your lifecycle support strategy - Development THROUGH End of Life. BTW - Contractors often offer GPR for free as part of their bid strategy, if they are requested in the RFP.
I am sick and tired of the incompetent political hacks running the acquisition commands - they have no clue what they are doing. When do we get professionals in charge that are measured on their performance and on how the systems they develop work?
"you have to be competent enough to ask for them up front"
DeleteIt's a little more difficult than just remembering to ask for the data rights. The Navy asks all the time. The problem is that many companies (most?) just don't want to give up proprietary data that provides them with a market advantage and that's perfectly understandable. The Navy understandably wants the data and industry, equally understandably, wants to hold tight to it. That's an almost unsolvable dilemma.
The only viable, long term solution is for the Navy to generate their own data via realistic testing.
Now, here's the big attraction to testing for the Navy: they will likely wind up with better, more accurate data. Let's face it, no manufacturer has their own private 'live' test range (depending on what the product is) so their data is likely incomplete or inaccurate, at least in part. So, with a little effort, the Navy would wind up with better data than the manufacturer has!
"When do we get professionals in charge"
When we make competence a promotable characteristic ... which it is not, at this time. Right now, gender sensitivity, diversity, wokeness, climate, etc. are all valued more by promotion boards than professional competence.
I'm pretty sure your question was rhetorical but I thought I'd throw the answer out there, anyway, just for fun !
CNO - You are right it was rhetorical and good rejoiner!
DeleteI am puzzled by your statement that the Navy asks for data rights and companies refuse to give them. If a company does not bid the rights the Navy requests, then the bid is non compliant and they lose out. I have never seen a company refuse to bid rights that were requested. They might try to bid a high price, but again they will lose. My experience has been that the Gov "forgets" to ask for them in the RFP. Do you have a case where a company refused to grant rights they bid?
"Do you have a case where a company refused to grant rights they bid?"
DeleteIf I understand the dimensions of your question correctly, no, I have no example of a company having bid/contracted to provide a product AND data rights and then refusing to provide the data. To the best of my knowledge, that has never happened and, if it did, it would trigger lawsuits and all manner of legal fights. That's not what's happening and not what the post is about.
What is happening is that the Navy is speaking with manufacturers prior to bids and contracts and finding out that the manufacturers won't supply data and, if it comes down to it, will simply not bid or not supply a product if single-sourced. In several well documented cases (see the post and DOT&E annual reports), the Navy has gone ahead and purchased the product without data rights and then moaned when the manufacturer did exactly as they warned and refused to provide data.
The Navy has a habit of sole-sourcing equipment, signing purchase contracts, and then trying to obtain the data rights after the fact.
The problem is not the contract writing, it's the industry refusal to enter into a contract that includes data rights.
What do you do when you've committed to a 55-ship LCS fleet, begun building them, and then you find out that the manufacturers of the radars you want to use (Seagiraffe and TRS-3D/4D) won't sign over their IP rights? That's the problem the Navy faces.
The combat management software suites are another product that the Navy is having trouble getting source code for.
The F-35 is another example, LM won't give up data rights to many aspects of the program and it's hurting the overall program performance.
The issue is not companies refusing to honor contracts and it is not the Navy forgetting to ask for data rights in the contract. The issue is that the Navy's product choices are severely limited if they exclude companies that won't provide data rights. A lot of good products can't be used by the Navy under those conditions and, in many cases, there aren't a lot of alternative products to choose from. The lack of choices is why single-sourcing is so common in military contracting.
Supposedly (I haven't been able to absolutely confirm this), the Navy wanted to use the Thales CAPTAS VDS but couldn't get the data rights and has now been forced to ask Raytheon or someone to build them a VDS, never having done so before. You know how things turn out when a company tries to build an advanced, highly complex product for the first time. It never turns out well. The entire LCS program is an example of that.
The Navy's refusal to conduct physical tests as a supposed cost-saving measure is severely impacting their product choices because they're finding out that no one wants to give them IP rights! The decision to forego physical testing to a large degree is very short-sighted and unwise.
" If a company does not bid the rights the Navy requests, then the bid is non compliant and they lose out."
DeleteThe Navy's practice of sole-sourcing severely limits the Navy's ability to declare bids non-compliant and toss them out. If you declare your sole-source non-compliant, then what do you do? By definition, there is no other supplier (source) you can turn to. The Navy has created their own problem by instituting various practices and philosophies that have resulted in a dramatic consolidation of the defense industry and created sole-source situations. To a large degree, it's a self-inflicted problem.
Another aspect of the problem is the use of foreign suppliers. While there are many good foreign products that can and should consider, our legal rights become quickly murky and problematic when dealing with a foreign based supplier. International legal disputes are notoriously difficult to successfully pursue and by the time they are resolved (if ever), the need has likely long since passed.
Your points are well made, I just have not seen this degree of sole source. The procurements that I have been involved in or close to: DDG-1000, LCS, MUUV, JSTARS ground stations, and numerous smaller ones all were competitve - at least at the beginning. I wasn't involved in the rights portions of the all of the bids, but the ones I did see asked for GPR. I still see this as mostly lousy acquisition planning and poor negotiating on the Navy's part.
ReplyDeleteEven in hte case where prior discussions with the vendors indicate they will not bid, well that is clever negotiating on the vendors part. When management has to make a decision to walk away from money, that seldom happens. More likely, the senior acquisition people are worried about their cushy second career opportunities.
As for SW rights for maintenance, you are right, the Navy does not have a clue, nor does it state in RFPs how and who will do the SW maintenance and getting the rights to support a SLCMP (if they even wrtie one). Again no focus on the Lifecycle of the product at the beginning.
I closely follow the DoD contract award announcements and I would estimate that around 20% of the contracts are sole-source.
DeleteThe 'bigger' the product and the more complex it is, the more likely it is to be sole-sourced. That makes sense. Run of the mill pumps, valves, chairs, and HVAC ducting are not unique and are readily available from multiple sources and no one cares about the data rights for a pump.
Radar, sensors, combat management software, missiles, naval guns, etc. are often (usually?) produced by only one company and that's where data rights are both important and jealously guarded by the manufacturer.
I don't know what types of products you were involved with in purchasing but certain types, as listed above, are generally sole-sourced.
Our carriers, by the way, are sole-sourced ! :) I know, that's not exactly what we're talking about. Just a humorous perspective.