An exploratory media piece on the dynamic visual identity of BBC Two.AI automation in media: concepts, data sources, automation architecture, and editorial standards.

«Variant 2”—an independent research media project on the evolution of the visual identity of the British channel BBC Two.

At the heart of the project is the iconic number 2, which served as the basis for the channel’s dynamic identity in 1991. Over the following decades, it changed materials, motion, and its interaction with space—from metallic and glass objects to animated characters and abstract digital compositions.

The media explores these transformations through the lens of communication design, demonstrating how persistent and variable elements create a cohesive visual system.

The main question of the project: how can an identity constantly evolve while maintaining brand recognition?

The history of graphic design is often presented through singular, iconic projects. We see finished logos, intros, and animations, but we rarely get the opportunity to systematically explore the principles that unify them into a cohesive visual system.

Furthermore, historical materials exist across various digital archives—they vary in descriptive depth and require individual research.

The goal of «Variant 2» is to consolidate BBC Two archival material into a sequential research media piece that explains the principles of dynamic identity through form, material, composition, and motion.

Automation powered by AI allows for the regular release of content based on pre-vetted historical data.

The primary audience for the media are students and emerging professionals aged 18–30 who are interested in communication design, branding, typography, motion design, and creative technologies.

The secondary audience includes practicing designers, art directors, and researchers of visual culture.

For this audience, the ability to not only find expressive references but also to understand the principles behind their creation is crucial: why a specific visual solution works, how movement affects the character of a sign, and what rules ensure brand recognition.

«Option 2» offers a transition from merely observing design to analyzing its structure.

The main media category is IDENT OF THE WEEK, which features short analyses of historical BBC Two idents. Each post includes the title and year of the ident, a verified historical context, a description of the visual technique, a design analysis, and a link to the original source. Additional sections expand on the research. MATERIAL STUDIES focuses on materials, textures, lighting, and the physical properties of objects. FORM & SYSTEM examines the constant and variable elements of an identity. MOTION ARCHIVE investigates movement, animation, and object interaction. DESIGN EVOLUTION discusses historical redesigns and technological advancements. OPEN LAB encourages the audience to experiment with visual systems independently. The primary platform for this media is Telegram. Seven main posts are scheduled each month: six automated pieces and one in-depth authorial study.

The primary source for this project is the BBC Motion Graphics Archive at Ravensbourne University in London. It contains historical television bumpers, descriptions of design decisions, information on creators, and production technologies.https://www.ravensbourne.ac.uk/bbc-motion-graphics-archiveA supplementary source is The Ident Gallery, an independent archive of television graphics. This is used to correlate historical periods, locate specific bumpers, and refine the chronology.https://theident.gallery/The third type of source consists of materials from design and animation studios involved in creating the bumpers. For example, Aardman publications allow an examination of production techniques and the creative approach to the BBC Two identity from 2018.https://www.aardman.com/short-form-commercials/bbc-two-ident-squish/Priority is given to primary archives and materials from the direct participants in the projects. Secondary sources are used for additional verification.

The foundation of this automation is a structured database of historical BBC Two intros.

Each object is described using a set of attributes: title, year of creation, historical period, creators, material, color, movement characteristics, the digit’s role in the composition, and production technology.

Additionally, the original source link, verified historical facts, the creator’s analytical observations, and usage rights information are preserved.

For example, the 1991 «Blade» intro could be described through the combination of a metallic material, a physical impact, and the recognizable shape of the digit «2.»

This structure allows not only for the automatic creation of publications but also for comparing historical objects based on unified visual characteristics.

Thus, the archive becomes a research database rather than merely a collection of images.

Before including material in the editorial database, the title of the feature, date, authorship, and historical context are verified. Each confirmed fact is accompanied by a source link. If the information is contradictory or insufficient, the material is sent for further review and is not cleared for automatic publication. Special attention is paid to copyright. BBC images and videos cannot be freely copied or republished solely based on their availability in the archive. Therefore, the basis of the visual content consists of original analytical diagrams, typographic compositions, and graphic reconstructions. Original historical works are accompanied by links to archives. Automatic image publication is only permitted if there is confirmed permission for its use.

Media production is based on the division between human and machine information processing.

In the first stage, a human researches archives, selects materials, verifies historical data, determines visual characteristics, and fills out the editorial database.

After the entry is approved, automated processing begins.

The system schedules the prepared material, feeds the data to the language model, and generates the publication according to the established editorial template.

Next, the text undergoes automatic verification, and the graphic service creates a visual card based on the ready-made design system.

After successfully passing all checks, the material is published on Telegram, and the system saves the issue details.

The editor monitors the results, analyzes errors, and updates the workflow rules.

The person is responsible for the research and substantive foundation of the media. Their tasks include selecting historical materials, verifying primary sources, clarifying dates and authorship, classifying archival footage, and formulating analytical observations. The editor determines which visual characteristics hold research significance and what conclusions can be drawn from them. The designer develops the media’s visual system: the grid, typography, color palette, graphic templates, and data presentation rules. The person also approves new entries for automatic publishing, reviews contentious material, and analyzes the quality of published content. Thus, historical accuracy, editorial stance, and visual identity remain under human control.

The automated system is responsible for repetitive stages of content production.

On a set schedule, it selects one of the prepared records and checks its readiness for publication.

A language model converts the confirmed information into a structured publication text: a headline, a description of the historical object, and a design commentary.

The next step is an automatic review of the text to ensure compliance with editorial standards and source data.

After this, a graphic service inserts the title, year, and visual characteristics into a pre-designed card template.

The finished materials are sent via Telegram, and the results are logged in the publication journal.

If an error is detected, the system halts the publication and notifies the editor.

For the technical implementation of the media, several interconnected services are intended to be used.Google Sheets functions as the editorial database. The spreadsheet stores vetted historical data, research categories, and material statuses.n8n manages the sequence of actions: it triggers the scenario, selects the data, passes it between services, validates the results, and handles errors.The OpenAI API is utilized for generating texts based on approved information.An SVG or HTML-based graphic templater automatically forms visual cards according to the design system.The Telegram Bot API ensures automatic publication of materials and provides technical confirmation of transmission.This architecture is designed for further implementation and testing.

The language model’s task is to transform a verified historical record into a short, precise, and substantive publication.

The text is created in Russian using a research-oriented, professional, yet accessible style. The recommended length for the main body of the text is 500 to 900 characters.

Each publication must include a feature title, the year, a description of one visual technique, a design insight, and a link to the original source.

The AI is prohibited from independently adding historical facts, inventing author names, dates, or production technologies, or using unverified judgments.

The model must rely exclusively on the provided data. If there is insufficient information to create a credible publication, it must report this and flag the material for manual review.

The primary quality criterion is factual accuracy and the presence of a specific design conclusion.

The visual language of «Variant 2» is built upon the aesthetics of a digital archive and research catalog.The core elements of the system include a modular grid, bold grotesque typography, a neutral dark background, and a greenish-cyan accent.All automatically generated cards use a unified template with a consistent layout for the title, image, and technical captions.Required parameters include the title of the piece, year, material, form type, and motion characteristic.The automation changes the field content but does not violate the established compositional rules.The graphic service uses only its own analytical imagery or pre-approved materials. Generating images that falsely resemble authentic BBC archival footage is prohibited.The recommended card format for Telegram is 1080 × 1080 pixels.

Before publication, every piece of content undergoes several stages of automated review. First, the system verifies the presence of a confirmed source, correct title, date, and required editorial fields. Next, the system cross-references the generated text with the approved record, checks the publication length, the inclusion of a design output, and the accuracy of the link. The licensing status of the used image is checked separately. Following this, the system prevents the duplicate publication of the same material and verifies the readiness of the graphic card. If all conditions are met, the publication is cleared for release. If even one mandatory check fails, the release is blocked, and the entry is routed to an editor. Automated verification does not replace historical expertise; therefore, new facts and disputed interpretations require human confirmation.