Intergalactic Summit

Some time in the past I had been asked about my decision networking studies, in specific my "dronehouse" simulated environment lab. Questions of transparency were brought up. To help assuage fears, I thought it best to release some of our findings and procedures from this study. If you are interested, please feel free to peruse the overview below, and should you wish more information, the full documents are available from the House of Records for a small fee. If you are a SAK alumni, you may also get the files from SAK Todaki campus where the files were registered.

My thanks to the SAK Todaki Information Science Department, to A. Vanamoinen, M. Abellia, S. K. Umotekke, A. Oto, and of course my dear colleagues at LDRBC.

- S

Analysis overview of Long-term Discrete Decision Network Simulated Environment Study
Project 'Dronehaus' preliminary findings report, public edition.
Lai Dai Research Biomedical and Cybernetic
Artificial Intelligence and Decision Networking Division
Research Associate Scherezad

For public release

Overview
The LTDDNSES (Dronehaus) project is an open-direction study of commercially-available decision network systems. A large number of commercially available drone control boards (DCBs) are networked with a computing cluster simulating a single non-real Constellation of space. Each DCB interfaces via a simulated drone of the appropriate type. The networks are divided into three roughly equal groups: 1) A control group which receives no commands, 2) A Simulated group which receives commands from a simulated group of Pilots, 3) A Real group which receives commands from a group of research associates. The behaviours of the DCBs are recorded and analysed for divergences in expected behaviour.

Study Components
Several discrete groups of resources are used in Dronehaus:

- Dronehaus Simulation Server (DSS). The DSS is a Kamkatkka 515 architecture server cluster hosting 1,024 quantum processing nodes controlled by a Lai Dai Information Systems network controller. Four battleship-grade ECM units have been repurposed to provide discrete emergency shutdown alongside the standard killswitch systems. The unit is powered by a repurposed Scorpion Battleship powerplant.

- DSS Environment Simulation (DES). The DES is a custom-built environmental simulation of a fictional Constellation. Ten Systems were created using random sampling of known Systems in the Cluster. The DES is populated by 120 simulated stations, each capable of spawning simple mining ships, security ships, and couriers. DES population activities are dictated by a semi-random schedule. System and Object names were chosen by SAK Todaki graduate students, and are named: Simard, Wakanna, Etebong, Uluetton, Umetton, Oriet, Terra, Alouette, Bixby, and Apa.

- Drone Control Boards (DCBs). DCBs used in this project come from three sources of providence: salvage, the open market, and construction efforts. Salvaged boards are inspected by LDRSS electrical and information engineers for network security before inclusion in the system. Purchased and constructed boards are likewise tested, but do not undergo a prolonged isolation period before inclusion. Dronehaus currently has 12,447 DCBs networked, with 1,806 Heavy, 4,290 Medium, and the remainder Light Drone controllers. 14% of the population are EW DCBs, and 19% are Logistical DCBs. Only 4% are Sentry DCBs.

- Dronehaus Control and Observation Interface (DCOI). The DCOI is the control room used to issue commands to and receive results from the DSS. It is a single room adjoining the DSS physical plant. It is typically staffed by five researchers, with no fewer than two and no more than twenty at any given time.

Procedure
All DCBs are first disconnected from any external hardware and inspected for network stability and health according to State Information Control standards. Any failing boards are retained for separate study in other projects. Boards which pass standards are included into a Kamkatkka processing bus which is, once synchronized, connected to the main network.

DCBs which are part of the control group are seeded in random groups of five and distributed at celestials throughout the simulated Constellation. These are monitored for activity, but are otherwise left alone. DCBs in the Simulated group are also seeded randomly as are the Control flights, but are also "patriated" to one of the simulated factions within the system. The faction in question enqueues a command for an available ship to visit the new drones and encorporate them into its structure. DCBs in the Real group are seeded randomly, but are assigned to one of three researcher-controlled factions. These are issued commands according to the research of the individual researcher.

Information about the state of each DCB is polled by the central processor of each Kamkatkka bus. Polling rates are non-determinant, and are based on activity levels within each DCB as well as a comparison of relative activities of the simulated factions in the case of the Simulated group.

DCBs controlling simulated drones which are "destroyed" go through a short diagnostic and refresh cycle before being refitted into a new simulated drone of the same type. After a given mission or flight, simulated drones are given a thorough inspection, cleaning, and repair. Control Drones are supplied with enough logistical drones within their flights to ensure that this can happen naturally and without interference.

(Continued)

This is fascinating. As a Gallente pilot I have a long history of training and working with drone flights, and the perhaps eccentric habit of imputing individual quirks to their behaviour.

My principle scientific background is not in computer networks or I'd be spamming your company with my CV already. However, I am a trained anthropologist who's published on cybernetic (in the social-philosophy sense, not the implant sense) communication and "self-organizing" social movements. So I suppose if you start seeing signs of learned group differentiation you might give me a call!

Are you planning on developing an Anoikis module for the simulation? The gravitic effects out here and disruptions to fluid router functioning have, I am beginning to suspect, some long-term effects on drone flight behaviour.

Absent that I shall simply wait with baited breath for further reports. Bravo.

Ms Okazon:

We have no plans at this time for including an Anoikis simulated environment, as we would go outside of scope on the project. However, it's a very interesting idea, and with enough interest I may be able to set up a second DSS cluster to include the environment. I suggest getting in touch with S. K. Umotekke at SAK Todaki IS Department if you would like to discuss including this in the project - I'm sure he would be very interested in the potential research!

You may also be interested in examining Project Valhalla as conducted by Pator Tech. They have identified three Control groups that are engaging in some unique signalling behaviours that have resulted in a strange rotating-flight-leader pattern. It's suboptimal for standard flight plans, and they're still trying to figure out what the change is about. My guess is that we're seeing a token passing behaviour, but I'm not sure.

Ms. Otsito:

You are of course welcome to observe and or participate! There are three levels of engagement. The first is through our Dronehauscam. Go to the SAK Todaki Galnet portal and search through the active research projects page for Project Dronehaus. One of the options in the descriptor page is Dronehauscam. It's currently focused on a flight of seventy Control Group drones that are engaged in perpetual combat with a neighbouring Simulated group. Many of the drones have been given names by visitors to the site, and there's a small but healthy betting pool as to expected lifespans. Project Valhalla is also monitoring this group very closely.

If you would like access to our data sets instead, you may do so for a nominal fee through SAK Todaki or the House of Records.

If you would like to conduct active research, visit our offices in Todaki and we can get you set up! there are a few NDAs to sign as well as some procedural details, but they are fairly simple, and are for everyones' mutual protection.

Thank you for your questions!

I'm outside my expertise here, so if I'm way off base say so, but I'm curious about this. Quarantine works in biological hot labs because a lot of them can't live outside a host. But what does isolation do for potential code-based threats?

Even though I'm not particularly tech savvy in this field, I have to say it's an interesting read.... What strikes me as a question, one that was perhaps answered, in which case it would require some underlining in the original text is; What are the benefits of the Dronehaus networks over standard drone control networks?