Gandalf Thumb Wrestles the Army

A few weeks ago several articles began bubbling up through the media on a memo and a report on the quality of Palantir vs. DCGS-A. This memo and report framed a debate. A debate that is ridiculous and I’ll explain why below.

Prior to proceeding, it should be said, I’m not here to sell either product, nor do I work for either company. I am going to let you decide.

What is DCGS-A?

Usually pronounced “deecigs” or phonetically spelled, the acronym stands for “Distributed Common Ground System-Army”. It’s a network built from the ground up by the Army with the intent (potentially) of replicating the architecture in the other services, thus the “A” for now, but possibly “N” and “AF” later. At the user level, it is a suite of tools. Not one tool, but many.

At the inception of DCGS, the Army needed a product that met several specifications in today’s operational environment. It had a system called ASAS (All Source Analysis System) that was passing into obsolescence due to the changing nature of warfare (an issue which is and has been debatable), the system functioned effectively (also up for debate) but was designed for a “force on force” type of warfare.

A strong point of the system was to be able to use various sensors and pick up battlefield ELINT, but even with that strong point, it had some serious gaps in functionality. To meet those gaps (among others) the Army hired contractors to build a “distributed” (think “cloud”) system that would allow a conventional or unconventional soldier be able to conduct battlefield fusion analysis usually at the Analysis Control Element (ACE) with just a laptop. Essentially the soldier would need a rugged, mobile system that would contain all the tools required to accomplish his tasks. This system also needed to be “plugged in” at a ground level to units sharing the battle space as well as a “reach back” component to talk to national level intelligence support (DIA, CIA, etc.).

What Does DCGS-A Do for the Analyst?

The DCGS GUI (Graphic User Interface) gives the analyst a list of tools to select from that are integrated and exportable to Microsoft Office (to make sure you earn that PowerPoint Ranger tab). Analysts get a mapping tool, a tool to interface with message traffic (and very important, export it), Google Earth, a social network analysis tool, a name variant tool, and various others. There are drawbacks, such as no real-time collaboration. The system does work in conjunction with various battlefield sensors and it has dramatically simplified a process that was measured in many more hours five years ago. The analyst then takes the message traffic, the mapping tool, and Microsoft Office suite and builds a product (usually a brief) to help the decision maker. I have summarized this extensively to get the concept of this system across, so yes there are features I have excluded for the purposes of this post. The key concept is that this is a mobile suite of tools that are a sub-component of a greater network infrastructure and they are designed to get the massive amount of intelligence collection and repositories to the hands of a soldier in the field.

What is Palantir?

Palantir is a magical artifact that Gandalf uses throughout the J.R.R. Tolkien saga. It’s more or less a crystal ball. If you look at Palantir’s logo, you’ll see what I mean. So what the heck do the two have in common? The creators of Palantir LOVE the “Lord of the Rings” saga. In fact, the FSR’s (Field Service Reps) will often have T-shirts that say “Save the Shire”, a comical reference to the home of Palantir in California.

Palantir was initially developed as a couple of locally created programs and protocols for PayPal to prevent credit card fraud (among other types of fraud). Realizing the utility of this, the creator’s then took the concept out on its own and developed Palantir (Finance). Several hedge fund managers bought it (and as I recall, so did the SEC, but don’t hold me to it). It rapidly grew in popularity and eventually caught the eye of the Agency. The Agency bought some licenses and found it equally useful. The marketing folk at Palantir realized there was utility in this program on a defense level and started pitching around the defense community. However, Palantir is expensive. An untested tool (except for testing by second and third hand sources) that is expensive, is an acquisitions manager’s nightmare. Some of the questions that must be answered are: Does it integrate into classified systems? Does it protect critical data on integration? Does it integrate with other baseline products? Most importantly, is it worth the price? Does it provide value added?

Palantir is one tool. It has subcomponents, but the underlying mechanism by which it works is extensible markup language (XML). You can create application programming interfaces (API) and integrate functions into it, but the API’s have to be developed. Additionally, the DoD has some unique architectures, so it’s not just a “code and make it happen” process. At classified levels, it gets even more difficult as you deal with things like encryption and access.

What Does Palantir Do for the Analyst?

Palantir is not an “Improvised Explosive Devices (IED) finding program”. Wow. What a colossal failure by the media. It doesn’t find anything. Much like DCGS, it uses existing sensor infrastructure and message traffic and integrates them into the Palantir user interface. It does however allow you to “crowd source” your work by putting it on the network for everyone to see and work on.

The program allows a user to “tag” entities within the documents (but not modify the documents themselves) and from the tagging generate them geophysically on a map, or in a histogram, or simply as a social network. The major drawback to this process is that the “tagging” is only as good as the analyst who is tagging. It still requires a degree of vetting to ensure that multiple entities (person, places, or things) aren’t tagged repeatedly or as different items when they are really one entity.

The links between entities need to be tagged and the networks need to be identified appropriately. It is definitely a useful program. In the picture below, you can see a list of the entities on the right. Those entities were “tagged”. They are not an automatically populated list. However, tagging is such a quick process it vastly reduces the amount of time required to create relationships and thus a conceptual map of a network. A network of IED’s for example.

Gandalf Thumb Wrestles the Army

Thus, you have two types of tools. The Army builds the first one. This means the Army said, “I want to see this and this, and this is the type of architecture I want”, and then they ask if Lockheed or Boeing can make it happen. Army engineers are certainly part of the process, but a lot of the nug coding and telecom work is done by companies.

The second is built for finance, with defense uses. It is built ground up by a company and being marketed and sold to the Army (and the other services). The company brings its architecture to the Army, it requires installation, but does not have to be built from the ground up.

Does Palantir have folks that lobby for it? Yes. Does DCGS-A have folks that lobby for it? Yes. Is Palantir expensive? Yes. Is it worth the cost? IMHO, Yes. Is DCGS-A expensive? Less than Palantir, but yes, it’s expensive. Is it worth the cost? Yes. Can you compare DCGS-A on a one-to-one level to Palantir? No. Each does different things and is better at different things. This means you are comparing a duck to a moose.

This argument arose at a policy and bean counter (comptroller) level. The debate was meant to give publicity (for whatever reason) to one or both of the programs and does not in any way address the issue of whether or not these tools are useful to the analyst and thus relevant to whether it saves soldiers’ lives. If it does, both sides are guilty of a degree of hypocrisy. Both of the systems have saved soldiers’ lives, because both save time. It’s a subjective argument and, personally, I as an analyst would not like to go to the battlefield without either of these systems. In fact, the more useful tools I have, the more lives I can help save.

How do you save money? Don’t spend nearly a quarter of your acquisition budget on a piece of flying hardware that doesn’t work effectively. My two cents.

I encourage questions (you can find the Commander’s Handbook online) and feedback. Finally, any errors are mine and mine alone.