A Battle Unseen

[Doc]

Spencer Space and Test Center. Spencer Suldanor
Report: Year 1 evaluation of Platform KMD-LDSS 3a
564011051135Z EOS
Not for public dissemination

Summary
This report covers the results of our field test of the Launch Detection Satellite System (referred to as "Platform"), created by partnership with KMD, KMI, IBSI and SSTC. The platform has been deployed on five geostatic satellites in Geosynchronous Terran Orbit, and was in place nearly 8 months before it was publicly announced in June of this year. The year long field test includes data from Egelion, Northern Seleya, Artania (for counterfactuals) and Dovani, and involved five separate satellites. The results of the test are robust across different weapons platforms, and the technology has received a B+ from graders.

History
The Platform was developed in the aftermath of the white phosphorus attacks on several Kalistani Cities in 5612. Planners were caught off guard by the surprise attacks and were dismayed when it was discovered that nobody in a position to warn Kalistan had any indication that there were intermediate range missiles incoming: Lodamun, who had previously been working on anti-missile technology, had not detected the launches, nor had they deployed counter measures to prevent the missiles which flew over their territory to strike Kalistan as it slept. It took the missiles launched from Likatonia between 27 and 32 minutes to strike Kalistan. Had Kalistan had even 20 of those minutes as a warning, many thousands might have been saved.

The result was that a cooperation between the Military Intel Section of IBSI, KMD, and KalMilInCo partnered with Spencer Space and Test Center to begin developing new satellite based launch detection that can at least give Kalistan and her partners a heads up when missiles were launched, as well as work out trajectories of those missiles to give an idea of possible targets. The idea behind the development of the platform was that it would provide early detection, as well as intelligence as to the location of capabilities, silo locations and intent, with the aim of saving lives within Kalistan and providing intel to our security partners.

In 5614, planners began working out how to deploy this platform onto Kalistan's existing satellite network and to augment it with specialized satellites designed for just this purpose. The initial outlay for the program was ₭7.5 Billion Imperial Rubles over the first ten years. The technology took about 19 years to work out and ended up costing more than ₭20 Billion. The first prototype platform was ready to build in 5633, but another two years of engineering issues delayed construction of a deployable version for two years, before the a working platform-exclusive satellite was launched. That satellite platform had a number of immediately evident flaws, mostly relating to false positives (it would pick up camp fires, nuclear reactors, and jet exhaust) leading to a failure to discern.

In 5636, it was decided to attempt to use current weather thermal imagine satellites which contain high definition digital cameras to scan the surface visually, thermally, and digitally, and while the module was produced and tested, new machine learning algorithms were developed to tell the difference between benign heat sources, and ones which should be of concern to security planners. The new platform was deployed in 5637 on board a standard geo-static thermal imaging satellite. Test results in Northern Seleya were positive, and benign sources which might have set off a false positive in the previous system were eradicated.

In 5639, the most recent Platform was launched which further refined the discernment algorithms, utilizing repurposed weather tracking and forecasting satellites. All false Positives were eliminated during the test, and the system only missed a scheduled test launch about 3% of the time, while picking up several actual launches which were not planned.

The program was not publicly announced until there was robust and reliable data from the 3a test, at which point the intention to build a global early warning network was announced below the fold in the Kalistani Paper.

We are confident that we are going down the correct path with this system and are ready once more to begin attempting to build a platform specific satellite which is dedicated specifically to this system and based on its parameters, which will add an important part to our national defense program.

Summary Test Results of the Platform v 3a
Test results from 56390605 til PD**

N1 (Actual Launches): 63
N2 (Negative Launch Drills): 45
Positive Detection of N1: 61 (96.825%)
Negative Detection of N1: 2
False detection of Launch (N1 or 2) (Discernment): 0
Negative Detection of N2: 45 scenarios (100%)

**Please see annex A for full spec test results of all 108 scenarios tested in the 3A test, per five satellites.

Areas for Improvement
Spencer test staff, working with IBSI have included areas that need improvement.

1. Higher definition cameras need to be included on Platform-Exclusive satellites. IBSI would ideally prefer 7-10 cm resolution, but Spencer believes that no less than 15cm resolution is essential. This includes both visual and thermal imaging.
   
2. More human training of the machine learning algorithm is absolutely necessary and should be constant.
   
3. Highly recommend more solid state HDD and memory capacity, as well as doubling core processing power and speed would be essential for more reliability. 97% was relatively easy, the last three percent will be exponentially more difficult.
   
4. The biggest flaw in the system is the human operator element. Currently the system is not integrated between those in the analysis center in Spencer and the different posts throughout the country which are the key link to provide warning to both military countermeasures and civilian authority who would be responsible for getting civilians to shelter. Relying on humans relays dramatically increases time of response. KMI, IBSI and the Regular Forces will be required to more deeply integrate the early warning signals we receive from the satellite, which are transmitted at light speed, to the various stakeholders who need to enact the policy for the system to be effective. This involves human training, and clock like response, above and beyond overcoming the technological hurdles.
   

Recommendations

• Continue testing 3a for another year with upgrades
  
• Continue building system 4 for launch sometime in 5641-42
  
• Begin developing integration of early warning systems throughout Kalistan and Ananto
  
• Continue working on developing effective and integrated countermeasures to neutralize the advantages that other nations clearly have at the moment with regard to missile technology, perhaps based on Lodamese technology, if they are willing to share it, otherwise, begin developing it here at home on their theory.
  
• Suggest integration with Terran based seismic detection, and land based long range radar facilities to enhance confidence in satellite data.
  
• While integration is coming on board, begin rigorous training for human operators on the early warning systems, so that they themselves become locked on with the developing EWS Platform and performed assigned tasks robustly.
  

Submitted, and Approved
Ingrid Frio
Director, KMD-LDSS 3a Program
Spencer STC