ID : 1432-654 TITLE : Automation and the Seven Worlds TITLE2 : A Brief History and Overview of Current Systems AUTHOR : Seven Swords Special Service AUTHOR2: Division of Documents & Records C DATE : 066/2350 CLASS : Top Secret GOLF INFORMATION IN THIS DOCUMENT IS CURRENT AND SENSITIVE.
The following is a glossary of terms either unique to this field, or with meanings in this context that are different from standard usage.
Automation is a field that is relatively young, yet it has met with more resistance than almost any other. It deals with the science and engineering of systems to automate tasks of all kinds, from the piloting of a starship through jumpspace, to the repetitive welding of particular parts on an assembly line. Throughout its short history, automation has been alternately hailed and despised, promoted and banned. This document is an attempt to provide a brief history of automation in the context of the Seven Worlds. In order to accomplish this, the history of automation in the Starguild will also be touched upon for comparison. Finally, some notes on the current research for the next generation of automation systems will be presented.
The Starguild, for all its vaunted technology and nearly limitless resources, is relatively behind in the field of automation. The Starguild does use computers and some expert systems, but they actively suppress and discourage development of robotics.
Robotics in the Starguild
The Starguild, since its very creation, has been opposed to robotics. The reasons for this are fairly simple: robotics are good at simple, repetitive tasks that are currently performed by Bondsmen; robotics require relatively sophisticated technology, making them difficult to employ securely in Bondsmen and Landcaste areas; and the use of robotics requires relatively technical training that would require much greater educational background in the Bondsmen and Landcaste. All of these points relate to the Starcaste's method of maintaining power: control of high technology, and the knowledge required to use, maintain, and create it.
Since the early days of the Starguild, the pursuit of research into robotics has been prohibited by Convention. Moreover, with the virtually limitless cheap labor of Bondsmen, there is no economic incentive to violate these Conventions. Hence, the Starguild has not only failed to investigate robotics, but has actively suppressed the basic knowledge to begin these investigations.
Recently, a new portion of Imperial Research & Development has been created to develop autonomous combat robots. The initial results of this (forbidden) effort are the Mantis combat robots. Although requiring strategic coordination by a human controller, the robots are a large first step. Once the tactical capabilities of the robot chassis are mated to a suitable expert system for strategic coordination, the Mantis may very well become among the most potent weapons yet fielded. It should be noted that despite official Imperial sanction, the Imperial Robotics Section is officially conducted outlawed research. This is unlikely to stop progress, however.
Other Automation
The prohibitions against robotics does not extend to computing or expert systems, however. The Starguild makes extensive use of general computers, for tasks ranging from scientific research to corporate accounting. Every citizen of the Starguild is able to use simple computer interfaces, although knowledge of programming and computer design are limited to the Starcaste and a very few Bondsmen and Landcaste (such as some on Rhand with TRIDENT/RMBK).
Computing power in the Starguild is cheap and pervasive. Virtually every device that can reasonably be computerized has been. Standard computer interfaces are either visual touch screens or voice control, with only a few examples of keyboard or pointer interfaces. These are restricted to the Landcaste military or Starcaste (science and military).
All known computing is based on photonic circuits, which can be implemented in a variety of materials. The low power consumption and high gate speed means that virtually any device can have the equivalent of a 20th century supercomputer inside. Consequently, the vast majority of people in the Starguild never realize how pervasive computers actually are; there are processors in any powered device, albeit with varying levels of complexity. The interaction between these processors is handled by a set of standard protocols, which include rudimentary encryption; it is known that the Imperial Secret Service, at least, has decryption keys for all standard consumer devices. Devices with sensitive information implement much tougher encryption, which is often proprietary to the Corporation that developed it.
Despite the enormous computing power present in any Starguild household, from Bondsmen appliances to Starcaste PCs, the Starguild has never pursued development of Expert Systems beyond a few special areas, most notably Jump drive computers and medical diagnosis systems. The Jump drive computer of a starship has been refined to an exceptional expert system, better than any known human pilot. However, most of this improvement was done in the early days of Jump drive development; there have been few major improvements since then (grav gate handling was one).
Medical diagnosis systems are the other area of Starguild expertise in expert systems. Several Corporations (and Subcons) make expert system diagnosis tools to augment the abilities of a physician. Scan, in particular, makes several models, from fixed installations that can detect and diagnosis any known ailment, to small packs for combat medics that are restricted to combat wounds and effects.
The lack of research into expert systems stems mostly from the limitless manpower of a Starguild corporation. There is little reason to expend effort to build an expert system that can replace 10 people, when you can simply recruit 10 more people for no effort. There are some situations, such a Jump Computers, where a human is just not reliable enough to be entrusted with a task. Similarly, there are expert systems to help medical personnel, where there is not a limitless supply of Starcaste personnel. Other situations also call for expert systems: communication routing, military signal intelligence, data mining of scientific and commercial data which cannot be released to the Landcaste or Bondsmen, shipboard diagnostic systems to improve the efficiency of repair crews, and so on. Most places that limited personnel are a problem, the Starcaste has developed some sort of expert system to augment the few people available.
The Seven Worlds, due to a chronic lack of personnel, has taken automation beyond anything the Starguild has ever dreamed. Without automation, the Seven Worlds would never have held out for so long, or been able to salvage anything from their destruction.
General Assemblers
Mycaenas, Inc. had always had an exploration arm that searched space for new worlds to exploit. Like other Starguild corporations, exploration was carried out by crewed ships that Jumped to systems likely to have useful planets, based on the analysis of remote sensing data. Also like other corporations, Mycaenas lost ships and crews. Not often, but enough that exploration crews were well-paid and treated as being slightly suicidal; these crews willingly Jumped to systems with unknown grav coordinates, beyond the range of any possible help, possibly into hostile situations, for the hope that a useful planet existed. Given the expense of a starship, and the difficulty recruiting competent crews, most corporations wished for ways to reduce the risk of exploration. Most corporations contented themselves by improving remote analysis, seeking to insure that exploration crews only went to useful systems, and knew what to expect when they arrived. Mycaenas did this as well, but they also pursued a different strategy.
Designers of exploration starships had long asked crews for feature wish lists, doing their best to implement them. In early 2089, a young engineer involved in the design of a new exploration ship began talking to some exploration crews, asking a new question: "why did your last mission end?". By looking at these responses, combined with the incident reports on file, the engineer reached several conclusions:
To the engineer, the obvious reaction is to design a ship that can build any parts it needs, using materials present in the target system. She presented her ideas to Frank Dietz in 2089, and was met with skeptical interest. Dietz saw merit in the findings and suggestion, but did not believe such a ship could be built. Regardless, Dietz authorized the engineer to recruit a few others and proceed with development on a limited budget.
This limited budget drove the design team to use existing technology and components anywhere possible. Thus, the initial attempts were to miniaturize existing production machines. After multiple failures at trying to build a machine for every part, yet still fit all of them into a starship hull, the design team decided to rethink the problem from the start. A computer programmer suggested that rather than a specialized, efficient, machine for every part, use a general, programmable, machine that could produce any single part.
By 2093, the prototype Generic Manufacturing and Assembling Device, or general assembler for short, was ready. This machine, built from modified standard CNC mills, could produce virtually any mechanical part in a starship. It was, however, far slower and more wasteful than a specialized machine for the same part. Equipped with a full library of ship parts, a single general assembler took only a few cubic meters of space, but could produce any mechanical part, given time and materials.
Upon presentation of the successful prototype general assembler, Mycaenas authorized a much larger research effort into producing machines that could produce any part of a starship from raw materials. Multiple design teams worked on various modular components of the system: raw material discovery, extraction, and processing; mechanical manufacturing; electronic/photonic manufacturing; and manufacturing control. By 2099, the teams had designed the first fully functional system, comprised of modular stages; one for mining and refining ores and rock into useful materials, one for machining mechanical parts, one for building circuits, and an expert system that could integrate, control, and test the output of the modules. The total system acquired the name "general assembler", while the older mechanical manufacturing-only systems were renamed "general manufacturing systems".
The exploration ship Farseeker was the first to incorporate the complete general assembler. On its first mission, the ship lost its Jump computer and backup due to a micro-meteorite puncture of the hull. Once the hull and atmosphere were stabilized, the ship put the general assembler to work, deploying the mining facilities on a nearby asteroid. After a week of work, the general assembler had produced a completely new Jump computer, and spare parts for the hull and other damaged systems. The Farseeker completed its mission and Jumped home without a single casualty.
Only a few ships were ever equipped with the general assembler; the system could almost double the cost of a starship, and most exploration missions were becoming safer due to the improved analysis of target systems. Due to fears of industrial sabotage, Mycaenas never publicized the incident with the Farseeker and only used their assembler-equipped ships for the deepest exploration. Of the 5 ships constructed with assemblers, none were lost in exploration.
When Hryken stepped down from Mycaenas, Dietz made sure to obtain a complete copy of the plans for the final versions of the general assembler; as Mycaenas had never revealed its existence publicly, the system was not patented. The plans could buy a large amount of goodwill with any other corporation, should the need arise. When Hryken began planning for the exile, the general assembler became an invaluable asset. Most of Hryken's capital ships were retrofitted with general assemblers.
When Hryken led the exile to the Seven Worlds, the general assemblers provided a safety net during transit, and a ready-made industrial base upon arrival. Automated mining and manufacturing were operational within months of arrival in the Seven Worlds, as the ship-board general assemblers replicated larger versions for the colonies.
These general assemblers became the basis for the Seven Worlds industrial base and research into automation. General assemblers were programmed to build more specialized plants for common necessities (such as bolts, tools, clothes, etc.) and the improvement and expansion of their possible products never stopped.
As the Seven Worlds established their colonies on the planets and moons of the system, they began to break apart the general assembler system into more disparate pieces. The mining and refining machinery, with the addition of an expert control system and some mobile machinery (robots, essentially), became automated mining outposts; shuttle crews could stop by to conduct repairs and pick up material every few weeks. The manufacturing components began to be specialized to increase efficiency.
Based on the automated mining and manufacturing, the industrial base of the Seven Worlds grew faster than expected. Industrial output was limited by the speed of the machines, not manpower to run them. For the tiny population of the Seven Worlds, this meant the difference between life and death in the new colonies.
With the coming of war to the Seven Worlds, the automated manufacturing systems meant that far more of the population could take part in the war without crippling industry. With the first hints of war, the Seven Worlds began scattering automated mining and manufacturing facilities throughout the system. This made it difficult for the Starguild to cripple industrial production with a few successful raids.
As the war turned against the Seven Worlds, the engineers began turning the automated systems in simple defensive installations. Mobile machines became light weapon platforms, and the fixed installations gained automated turrets. The Starguild quickly started bombarding these installations, rather than expend soldiers trying to capture them. Although the Starguild military saw some value in these automated systems, they performed so poorly in combat that no mention of them was made to the Imperium; Seven Worlds Power Armor and starships were a much greater threat.
As the Seven Worlds were destroyed, much of the Seven Worlds' starfleet escaped. Within this fleet were ships with full general assemblers, and the plans for the more efficient, specialized assemblers developed over the preceding centuries. These assemblers built the industrial base for the implementation of Operation: Phoenix.
Expert Systems/AI
Without the expert system to control the components, the impressive general assemblers are little more than a collection of sophisticated milling machines. It is the controlling expert system that makes the assembler so valuable; it has the design and production information for virtually any item stored in its memory. Equipped with sensors and testing equipment, the expert system can replace a trained technician in checking manufactured parts for a defect.
Expert systems in the Seven worlds began with the control systems for general assemblers, but evolved into much more. The techniques to develop the general assemblers pushed the boundaries of automation design, even before the completion of the first general assembler. Hryken brought many expert system/algorithm developers to the Seven Worlds, and they continued their improvements. Most notably among their achievments was the integration of self-diagnosis and auto-repair capabilities; critical computing automation could include both redundant circuits and also generic circuits that could be programmed to replace any other circuit.
In addition to controlling manufacturing, expert systems were available for everything from medical diagnosis and treatment to repairing appliances. The Seven Worlds military took advantage of these improvements, creating expert systems to analyze signal intelligence, predict likely Starguild military maneuvers, and devise attacks against Starguild automation systems.
As TRIDENT took over the remnants of the Seven Worlds, they searched the ruins of the automated mining and manufacturing outposts. Some had intact backups of the control system. What TRIDENT eventually recovered astounded them. They had never seen such advanced expert systems, and immediately began to research them. The research was conducted at TRIDENT's most reclusive research facility, RMBK on Rhand. These expert systems, while not truly intelligent, represented levels of competence the scientists of RMBK had never dreamed possible. They began to stretch the systems to reach the improbable; full artificial intelligence. Sentient AI was achieved within the confines of an advanced BiComp, designated Model 9. TRIDENT also investigated the self-repair capabilities of Seven Worlds automation, incorporating their findings into such things as the ORCA.
With the military conflict on Rhand, the advancements of RMBK have been lost to the Starguild, although backups exist in the S4 archives. Assuming the success of Operation: Phoenix, research and development can continue where it stopped.
Since the beginning of automation, there have been those who sought to exploit it as a weakness in an enemy. From the first attacks against the factories of an enemy, to the sophisticated viral assaults against expert systems, all types of automation attack attempt to exploit the repetitive, perfect nature of the automation.
There are essentially two types of automation attacks: physical destruction, and algorithm hacking. The former type is the simplest; the physical mechanism of the automation is assaulted, with the intent of destroying or crippling the system. When successful, such an attack requires the automation system to be rebuilt from scratch. However, they are rather harsh and all-or-nothing attacks; it is impossible to destroy only one program on a PC by obliterating the processor, for example - everything is destroyed. Physical assault against the automation system can be very effective against systems that are virtually impervious to algorithm attacks, as these systems tend to be relatively fragile. Physical assaults against most military systems is difficult, as they are specifically designed to withstand enormous abuse.
The second class of attacks are known as algorithm hacks. These are attacks that exploit some flaw in an algorithm or implementation. The ancient "buffer overflow" attacks against computers are perfect examples of this sort of assault. Most algorithm hacks stem from a flaw in the implementation of some protocol or algorithm; by feeding the algorithm carefully constructed input, the attacker can cause the target to perform an action. Some attacks are purely destructive, desiring only to cause the target to crash or halt. Others are more insidious, and are designed to give access to protected data, corrupt trusted data, etc.
Algorithm hacks are difficult to develop, although often relatively easy to defend against. In the case of complicated expert systems, which often include automated defenses, it may require another expert system to find an exploitable algorithm hack. This has led, in some cases, to a small, quiet, arms race between target and attacker expert systems developing defenses and attacks. There are some expert systems which have been developed specifically to assault other computer automation; important systems connected to any world-accessible network have equivalent automated defenses that look for intrusion attempts. The most vicious of these defense systems launch counter-attacks against a detected attacker!
As with any static document, this section will be outdated almost as soon as it is written. Still, as of early 2350, the following are the known avenues of automation research in known space.
END DOCUMENT