[REDACTED] Military Intelligence Department
Submitted
[REDACTED] European Side Military Intelligence Department Copy Number: 6
[REDACTED] Asian Side Military Intelligence Department Copy Number: 7
No.: TS-[REDACTED]SAD00*
Security level: Confidential
Access authorization: Colonel and above
----------------------------------------------------------------
The years between 2045 and 2061 were some of the darkest and most oppressive periods of human history. The eruption of World War III caused a violent turbulence that destroyed the world's economic systems and production capabilities. The mobilization of armies and the breakout of war caused the resource of "humans" to dwindle. The spread of radiation also led to a constant reduction of inhabitable lands, which combined with various other factors caused humanity's reproduction rate to greatly drop. "Humans" became an increasingly valuable and expensive resource, as even the military industries which stood to gain the most from war became unable to afford the manpower costs to continue production. To reduce costs and retain their production power, several important military organizations began to research bionic humanoids(dolls) to replace their living workforce.
[ ] Production Type Frame-Robot Era
The first generation of production type dolls are technically not dolls, rather they are more suitable to be called androids - the design emphasized ease of production, therefore the early models tend to be stranger. To reduce labor costs and production time, Leibert General Manufacturing began developing an assembly line production method for frame-robots in the year 2033. The design of a frame-robot includes: a reliable structure that can handle most labor-intensive work; long automated operation periods, with 4 hours of connected operation and up to 6 hours of wireless operation; high versatility, able to swap parts based on the work specifications to reduce maintenance costs; corresponding AI system which can change jobs through switching internal software. Such a target was epochal at the time, which greatly troubled the development head Weitkin when receiving the order.
The first prototype of the production type frame-robots was completed after 2 years, internally labelled as ALR-50(T) [footnote-1]. The production costs of the prototype was extremely high, especially the installation time which required 130 man hours, to the point where it couldn't break even with its operation lifespan. Weitkin's group quickly made improvements on the prototype, and developed the two successive models ALR-50E2(T) and ALR-51C(T).
After extensive testing, the ALR-50E2(T) model was able to reach an average operation period of 20 hours per day, its endurance and recharging capabilities meeting the requirements as well. Its modularized design was also able to cut down on maintenance costs, however due to its relatively simple structure, it was unable to change jobs on the fly as required. Due to this, the Weitkin group developed the ALR-51C(T) prototype to test the results of combining a generalized design and a smart AI. The final test results of ALR-51C(T) was not ideal, its complexity causing the model to fall greatly behind ALR-50E2(T) during extensive testing, as well as the fact that there was no suitable AI system to assist in ALR-51C(T)'s testing, causing it to achieve subpar results in the flexibility tests. In the end, as expected Leibert GM chose the ALR-50E2(T) to become the mass production model, and labelled the final improved model as ALR-52P. On the other hand, the ALR-51C(T) model was put into storage, with no plans to further development. As the project leader, Weitkin understood that ALR-51C(T) possessed a great potential, as long as it could be improved and a suitable AI developed, the usability of ALR-51C(T) could break through the current technology ceiling. However, as the world situation became increasingly tense and the market gradually diminished, Leibert GM did not possess the resources to support two models of frame-robots. Since ALR-52P basically met the given requirements, and the improved model had greater performance as well, in the end even Weitkin was transferred into the development team of the ALR-52P series, while the ALR-51C(T) was slowly forgotten.
The next time ALR-51C(T) returned to the spotlight was 15 months later. Worcester United Electronics Manufacturing Company received an order from the Pan-Europe Military Union for the development of a series type automatic operation system. The system was mainly to provide an automated operation service for the "Assault Artillery" series vehicles, and its development was undertaken by the technical director of the company's sixth internal development lab, Mikhachi. To ensure the versatility of the system, the development team planned to carry out tests on a small-scale vehicle to obtain data. At first, they were preparing to create a test vehicle themselves, however initial planning revealed that the costs would greatly exceed their expectations, forcing them to look for a suitable test vehicle externally. The process of looking for a suitable test vehicle did not go smoothly at first as their technical requirements were too high, and such a vehicle did not even exist on the market. Just when Mikhachi was about to give up on the project, he encountered Weitkin during a chance encounter at a dinner party, who had risen to the rank of Robotics Director by then. The two hit it off immediately, and Weitkin was happy to provide technical support for the test vehicle for Mikhachi's automatic operation system. Under Weitkin's support, Mikhachi's development team was able to purchase the complete documentation and almost all the test data for ALR-51C(T) at a very low price. The installation of the automatic operation system was greatly accelerated with the help of the ALR-51C(T) model, and an important breakthrough was made in a short 2 months.
Due to the flexibility of the ALR-51C(T)'s skeleton, it was able to complete system commands with great efficiency. To improve its operation, the technical team did a big change on the overall structure, to make it closer to a human's body. The modified ALR-51C(T) was relabelled as TD-01, and the development team nicknamed it "Baby". Through swapping parts, TD-01 [footnote-2] could simulate different models of vehicle operation modes, which helped the series type automatic operation system to be completed ahead of time, and the final result pleased the Military as well. The success of the system greatly reduced the operation costs of the "Assault Artillery" series vehicles and increased the command efficiency of the Pan-Europe Military Union's mechanized troops, and the large amount of orders for the system led to a huge surge of profit for the Worcester United Electronics Manufacturing Company, as well as boosting Mikhachi's fame. However, Mikhachi did not stop at this automatic system, he keenly recognized that if the system could be improved into an automated AI to be used on TD-01, this technical breakthrough would outshadow by far the system that he provided to the Military. With this, he quickly assembled a development team of his own, and came to an agreement with Leibert General Manufacturing to let Weitkin and his own team continue the development of TD-01 and a suitable automated AI.
While developing the automatic operation system, Mikhachi and Weitkin had forged a strong friendship, and the heavy monetary investment in TD-01 also excited Weitkin greatly, as he began harboring hopes of a new version of the dolls. Mikhachi and Weitkin's teams merged together, and they moved into a research block equipped with machine tools and multi-stage computing mainframes. After half a year of research, TD-01's structure was strengthened and its software improved, as well as the experimental inclusion of small amounts of man-made muscles to support its movements. The newly enhanced TD-01 was renamed to CSD-02 [footnote-3]. During the development process, tests were being carried out in parallel and Mikhachi's AI development team continually made code adjustments and variable changes based on the test data. Based on the test data, CSD-02 and its AI's compatibility kept increasing, however the problem of its structure being too complex was still unresolved. Due to this, Weitkin was forced to refer to ALR-52P's design parameters to try and reduce CSD-02's parts requirements.
Besides the military, several civilian enterprises had become interested in the project. Before long, some companies had started buying licenses for the IP rights to CSD-02's system and attempted their own enhancements. Among these, one of the most successful was Szchelber Advanced Technologies. From the reports on the AI tests, Szchelber realised the potential of the dolls and contacted Worcester to obtain the rights to research and develop the CSD-02 platform. During the development of the AI of the AA series vehicles, it was proven that the AIs could not completely replace manned systems, thus the tests on manned systems had never stopped. However, initial experiments with manned systems had caused fatalities before, so the military had been facing trouble with obtaining sufficient data since Worcerster had turned to fully focusing on AI development. Because of this, Szchelber which had obtained the rights to the CSD-02 platform received a request from the military to develop a specialised doll for testing the AA system load. The doll developed by Szhelber had increased force load, and gravitational sensors were installed in important areas so that data on the doll's movements and acting forces could be obtained directly, helping the military to carry out tests and adjustments. The first set of sensor systems was completed in the year 2041, the doll number being CSD-03H, and this set of sensor systems continued to be used in successive models of dolls, greatly increasing their agility and accuracy of their movements. Due to the success of the test doll, the military increased their attention towards this technology, and under the cooperation of both sides contacted Worcester, hoping that the cooperation of the military and the two companies would result in the development of a doll system that could truly be used on the battlefield.
As the cooperation of the development company and the military was rather successful, the new doll models and research on their AI quickly gained the attention of some higher ranked military commanders. They thought that if the research was successful, perhaps mechanical soldiers could be developed, which happened to be a great temptation for the military which lacked manpower severely. Among these, General Carter of the Army's General Staff was unusually interested in the dolls, he believed that if the dolls could reach a certain technical level, they could replace the infantry completely, greatly reducing casualties and increasing individual firepower. To this end, he persuaded the allocation of special funds and set up an internal committee to supervise the cooperation of the two companies. To ensure that the research project would proceed more smoothly, both companies invested money to establish a new company specializing in dolls, in order to build more professional combat dolls for the military. This company was named "Important Operation Prototype Manufacturing", later known as I.O.P.
Footnote-1: ALR refers to Auto Labor Robot, C refers to Common, T refers to Testing.
Footnote-2: TD refers to Testing Doll.
Footnote-3: CSD refers to Common Strengthen Doll.
[ ] First generation T-doll Era
Starting from the preparation works to the official establishment of IOP took three years, by then it was the year 2046, and World War III had already erupted.
The establishment of IOP took a considerably long time, due to possessing huge funds and having the support of the military, Weitkin did not immediately invest all his energy into the development of new dolls, instead carefully searched for a partner. From Weitkin's point of view, the future of dolls was vast, however it also meant that the technical problems faced would be even greater from then on. In Weitkin's strategies, besides fulfilling the military's combat requirements, civilian services and security was also a huge potential market. To expand their reach globally, IOP registered three wholly owned subsidiaries in America, Ukraine and China respectively. Among them, the subsidiary in Ukraine often had business dealings with the then popular Sangvis Ferri Manufacturing. Unexpectedly, Sangvis Ferri was far more interested in the ALR-52P model rather than the CSD-03H prototype - due to eastern European traditions, they placed more importance in the structural integrity and durability of the dolls. Though it was somewhat different from Weitkin's original plans, however due to the large investment from Sangvis Ferri, there was little reason for Weitkin to refuse; Sangvis Ferri was also willing to send in a team to work with IOP in researching automated AI systems for the dolls, to him it could not be better. However, this ideal situation did not last for long, as World War III erupted in the year 2045 and the military's requirements for T-dolls greatly increased, and Weitkin was forced to immediately establish IOP and begin accepting doll orders from the military.
Due to high production requirements, IOP outsourced some of the contracts to Sangvis Ferri and transferred some technical documentation to them so that they could carry out improvements on the T-dolls. At the time, the AI of the T-dolls was still comparatively weaker, and they were unable to effectively become the main battle force, instead being mostly used as fixed guards for the military police in the reserve areas. Due to the complex structure of the CSD series T-dolls, they frequently suffered breakdowns, instead it was the "Jaeger" series developed from the ALR-52P by Sangvis Ferri that won the praise of the frontline. Officers and men on the frontline scathingly gave the CSD T-dolls the nickname of "live targets" to criticize their poor performance in battle. Because of this, the military began transferring their orders to Sangvis Ferri instead, as IOP's products became less valued. This greatly enraged Weitkin, and the relationship between IOP and Sangvis Ferri quickly soured. Even though they received much less orders due to the problem of reliability, Weitkin still believed that a design focusing on agility was the way to the future, and he spent the entirety of World War III trying to enhance their T-doll's AI systems and structure, but in the end there was nearly no results, which demoralised him and his team.
In 2049, the situation changed. Mikhachi's student spotted a technical team "90wish" on the Internet. They had published some articles on T-doll technology, and after reading some of the articles, Mikhachi felt that there was some prospect to it, and sent it to Weitkin. After reading the articles, Weitkin immediately recognized that 90wish's technical ability and developmental concepts were one of a kind, and began attempting to contact 90wish's members through multiple avenues. However, after 90wish published the articles, they had seemingly vanished off the face of the earth, and there was no further information about them anywhere. Reports from the Russian Ministry of Internal Affairs Network Security Agency showed that the publishing of the articles had not obtained 90wish's internal consensus, and the group had split apart and now held mutual hatred towards each other. Through his friend at the Ministry of Internal Affairs, Weitkin obtained information on two former members of 90wish whose lives were now being threatened, and spent a large amount of money hiring a private military contractor to protect them. After losing more than half of the escort members, former 90wish members Persica and Lyco were successfully sent to a safe house belonging to IOP. After some negotiation, Persica and Lyco agreed to work with Weitkin, and to hide their traces, Weitkin used a false identity to set up a small company called "Hermit", and used the method of miscellaneous business to carry out financial dealings with IOP. Hermit provided IOP with many advanced technologies, and using these technologies, IOP was finally able to complete the successor models to the CSD series in 2051 [footnote-4], and after enhancing and producing new combat T-dolls, was able to get back the military's T-doll orders from Sangvis Ferri.
The successive models of CSD later participated in many battles in the Nordcourt ground battles. Although the T-dolls' AI capabilities were still lacking, in areas where manpower was lacking, soldiers could use several first generation T-dolls to carry out surveillance and simple counterattacks. T-dolls took the place of soldiers in assaults and casualties, helping to reduce the loss of lives. Based on the feedback from battle, in return for sacrificing part of the CSD series' agility, IOP reduced the number of parts and increased the stability of the structure, as well as greatly increasing the density of the bulletproof layer and the titanium alloy skeleton. The improved model was labelled ACD-50/51/R/T military model. The ACD-50 model decreased the size of the head to reduce the chances of being hit, and the camera was covered in 10mm steel plates to avoid damage from fragments. The visual control and guidance modules were moved to the more flexible abdomen, while the signaler and processor was moved to the chest area protected by steel plates and connected with wave guides. Additional line control interfaces were added at the neck and tail vertebrae areas, so that in the case of the AI system being disrupted or destroyed, soldiers could continue to control the T-dolls using line control. The ACD-50 model's scanning and combat radius was 220 meters, with a single charge giving 28 hours of activity and 55 hours of standby time, as well as having low purchase and maintenance costs. The ACD-51 model was produced in 2051, with the main improvements being a lowered center of gravity so that it became easier to use heavy weaponry, as well as increasing the accommodation range and load in order to solve the problem of ammunition storage. The scanning and combat radius was 200 meters, with a single charge giving 21 hours of activity and 48 hours of standby time. These numbers were slightly lower than that of the ACD-50, however considering the overall increase in capability, the military didn't mind these small details. Other than use in combat, soldiers sometimes used the T-dolls for moving heavy loads, and due to the low rate of malfunctions and durable structure obtained good reviews. The R and T models were mainly used for scouting and special operation purposes, so the production volume was much smaller and they were not widely used. The ACD series T-dolls totalled 11730 dolls produced, among which around 2300 dolls were produced by Sangvis Ferri, and they remained as the most widely produced and used T-doll model in World War III.
Footnote-4: The main production models were CSD-08A, CSD-08Bs and the derivative IAD series. The IAD series later became the core internal module prototype models for the second generation T-dolls.
[ ] Second generation T-doll Era
One of the main factors for the growth of the civilian market was the change in the world situation. By 2051, World War III officially ended and the need for manpower in production and service areas as well as reliance on private military contractors steadily increased. Several larger-scale PMCs and transnational enterprises had direct business dealings with IOP, and among them G&K [footnote-5] was one of the more important clients. The founder of G&K, Berezovich Kryuger was originally the Ministry of Internal Affairs tactical commander who escorted Persica to IOP. After receiving the payment from that operation, he left the Ministry of Internal Affairs and used the money to set up his own company in 2053. G&K possessed a strong relationship with IOP, and since the cost of purchasing a civilian T-doll was much cheaper than hiring human soldiers, thus Kryuger was quite keen on the usage of T-dolls. Besides that, Kryuger also maintained a cordial private relationship with Weitkin, which resulted in IOP allowing Kryuger to have direct contact with Hermit to provide the latest data feedbacks.
Persica obtained a large amount of data from G&K, and using these she improved and developed the CSD-06/07/08 and SST-02A/03 models, as well as the Service Doll series which was specially developed for the civilian market. The Service Doll initial models were equipped with emotion simulation software and distinct genders in their outer appearance, which helped in the gradual acceptance of the dolls by the public. The successive models also possessed beautiful outer appearances, good communication experience and customizable service, which won them great acclaim. A large number of purchases from the government, public enterprises, service agencies and some private orders contributed to a great amount of income for IOP. After building up a series of technologies, Persica came up with the "Second Generation T-doll" conceptual theory. Persica's research direction obtained the support of Weitkin as better T-dolls could bring in more orders for IOP, however Lyco disagreed with this line of thought. Lyco's specialty lay more in AI research than T-dolls, and he believed that instead of updating existing technologies and networks, humanity's future lay in the research and development of AI technologies. Lyco hoped for more funds to carry out more in-depth research into T-doll AIs, however his application was rejected, which greatly demoralized him and resulted in his passion towards his work gradually decreasing. Slowly he began wanting to be free of Persica and IOP, and he secretly contacted Sangvis Ferri in many business meetings. Sangvis Ferri was glad to accept a former 90wish member, however Lyco was afraid that he would be targeted by IOP in revenge, and so Sangvis Ferri secretly contacted Kryuger and requested G&K to provide firepower support when escorting Lyco to Sangvis Ferri. In return, Kryuger requested for the research results of Lyco's subsequent research as remuneration for the escort, and offered to provide continual protection for Lyco at the Sangvis Ferri factory he would reside in thereafter. Sangvis Ferri and Kryuger quickly came to an agreement, and Lyco was safely escorted to Sangvis Ferri's underground research lab. Based on the agreement, Lyco quickly built a new factory and immersed himself into research and development, and after three years managed to hugely improve the capabilities of Sangvis Ferri's ALR series T-dolls. IOP quickly discovered that Lyco had secretly left Hermit, but considering their rival's power and their own patron's intervention, Weitkin was unable to take action. To maintain his own product's competitiveness, Weitkin decided to greatly increase Persica's funding, which directly contributed to the creation of 16Lab.
After Lyco left Hermit, Persica begin to notice the importance of software and networking for T-dolls. She and her team switched their emphasis to the development of etching technology. After confirming that 90wish would not be threatened anymore, Persica finally revealed her identity in the T-doll industry. After accepting IOP's funding, Persica began recruiting members, and in the year 2057 officially set up a technology research centre known as "16Lab". Having sufficient manpower, Persica quickly completed the "ASST" etching technology, which allowed T-dolls to greatly increase their proficiency with weapons and tools using specially compiled code. The etching technology was an open-source technology, and developers could download the engine and use it to write custom programs for dolls based on their needs. Persica appointed a team to carry out long term maintenance and updates on the etching technology, and using this she made suitable adjustments to IOP's dolls, which gave IOP an advantage on hardware-software compatibility against their competitors for a long time. On the basis of the Zener Agreement, she rewrote the dolls' mobile network docking structure. In the absence of human command, such dolls possessed limited group ability to act, and doll teams could be set up beforehand and act on orders sent through the network. These teams' action modes included following plans, scouting, analyzing and synchronizing. On the basis of etching theory, Persica prioritized researching a command network for T-dolls, as a complete network structure would allow human commanders to just give orders to a single T-doll in order to command the entire T-doll squad. After testing, it was determined that the automatic analysis and communication system could increase T-doll effectiveness by 37%, 16Lab named this new T-doll network command system as "Dummy network". While perfecting the command network, Persica also began design classifications on dolls of the same model, with hardware as the standard for differentiating the dolls into main dolls and dummy dolls. The main processors and synchronization hardware would only be installed in the main dolls, whereas dummy dolls would have less hardware equipment, with only the etching and operating software installed. Although this increased the production costs of the T-doll main bodies, however at the same time it decreased the production costs of the dummy T-doll bodies, thus as more T-dolls were bought for a single squad, the purchase price would further drop, further strengthening the price competitiveness of IOP's products. The T-dolls produced after 2057 all underwent modularized standard production, with etching software ports and dummy network systems built-in. Persica then defined the modular and standardized T-dolls as the second generation T-dolls, and published the complete technical specifications for them, with the intention of unifying T-doll operating systems and production standards.
In the same period, Lyco had his own ideas as to T-doll AI operations and modularization. His AI "Elisa" which used self-looping algorithms had already begun to take shape during the war, and the first generation T-doll command system relying on Elisa which was based on a tree command structure showed great performance when under direct command, with the encrypted mode and algorithms also being much better. Unlike Persica who used software to ensure compatibility with general weaponry, by utilizing Sangvis Ferri's strong production power, Lyco developed small-scale particle weapons with greater firepower. Before the design was finalized, he installed the particle weapons on the T-dolls to carry out tests, and performed adjustments and enhancements on the T-dolls based on the weapons. Although Sangvis Ferri T-dolls could not adapt to different weapons using software like IOP T-dolls, their unified designs allowed them to perform better individually. However, after the "Butterfly Incident" in 2061, Sangvis Ferri's research was forcibly stopped, and the related details were sealed in Area S09. As Sangvis Ferri's products stagnated, IOP basically controlled the main market for dolls. Other than providing combat T-dolls for the military and militias, IOP also developed many non-armed models for civilian usage. After the unification of the second generation T-doll standards, Persica continued making improvements to the modular structure of the T-dolls. To better control the T-dolls' safety measures, she integrated and locked all the combat-related data into a control core. The control core possessed many connections and ports, with a high compatibility with etching technology. The software would be directly installed in the control core, and if a T-doll's combat capability needed to be removed, just removing the T-doll's control core would achieve it.
After the year 2062, IOP essentially completed the standardization of the modules for combat and civilian dolls, and other than the ACD series that was specially modified for military purposes, the other non-military dolls were reduced to six models, which were the SST-05/05A/05A2 and SSD-62D/F/G models. The SST was one of the more qualified of IOP's brands, and the SST-05 series were IOP's flagship products, possessing high operation speeds, good weapon compatibility and higher storage space for the emotion simulation to make interactions with humans more natural. As one of the early special tactics models, the 05 model had slightly worse performance than the succeeding models, but due to its high production amount and bigger database for mental maps, this model had access to a great volume of combat data. For the 05A model, as an upgrade to the special tactics model, it could use data from the 05 model mental map database to increase its own efficiency, and besides requiring less software to be installed, there was also some improvement to its hardware [footnote-6]. The 05A2 was a special combat model, and its capabilities was closer to the military models at the time, with a much higher price tag as well. Since this model had less requirements for human interaction, IOP restricted its emotion simulation storage in order to install more software. Something of note is that the 05A2's mental data was individually encrypted and stored, allowing this model of T-doll to be safer when carrying out secret missions. On the other hand, the SSD series was a brand that was developed by IOP with both the civilian-use and combat-use markets in mind. In order to better integrate into human society, the SSD series have a large amount of storage dedicated to carrying out emotion simulation calculations, along with the installation of a third-party interaction system so that the average citizen would be able to use it brainlessly. Other than that, the highlight of the SSD series was the customizable outer appearance, with the D model being the service type and the F model being the ceremonial type. Both models had several hundred changeable outer appearance modules available for purchase, which essentially fulfilled civilian requirements towards service type dolls; at the same time, by installing a control core on the dolls when needed, the SSD model dolls also had an above average performance in battle, being able to fulfill bodyguard and policing duties which required limited destructive power. The G model were a special service model, specifically used for special duties above the police and below the military, compared to the D/F models they were much more expensive even though they had better capabilities. Due to their lower production amounts and the cost-to-value problem, they weren't seen much on the market.
Footnote-5: The full name is Griffon and Kryuger Military Contractor and Security Consultancy Services.
Footnote-6: G&K was extremely pleased with the 05A model, and most of the frontline T-dolls used by them are of this model.