AI-based system for creating dynamic poems based on IoT environmental data
The AI-based poetry generation system addresses the challenge of creating contextually relevant poems by using IoT environmental data to interpret and generate site-specific poems in real-time, offering an immersive and emotionally resonant experience.
Patent Information
- Application Number
- DE202024106336
- Authority / Receiving Office
- DE · DE
- Patent Type
- Utility models
- Current Assignee / Owner
- Filing Date
- 2024-11-06
- Publication Date
- 2024-12-24
- Estimated Expiration
- 2034-11-30
AI Technical Summary
Existing poetry generation systems lack the ability to dynamically generate contextually relevant and emotionally resonant poems in real-time using IoT environmental data, failing to integrate environmental variables like temperature, humidity, and noise levels effectively.
An AI-based poetry generation system that utilizes an environmental sensor network, a data processing unit, and a poetry generation engine to interpret real-time environmental data, incorporating mood, theme, and rhythm parameters to create site-specific poems through an LSTM model, with a vocabulary and syntax database, and output interfaces for display or audio.
The system generates contextually appropriate and emotionally impactful poems in real-time, providing a dynamic and immersive experience by integrating environmental data with linguistic versatility, enhancing user interaction and awareness of surroundings.
Smart Images
Figure 00000000_0000_ABST
Abstract
Description
Field of invention:
[0001] The present invention relates to the interactive arts of literature and environmental technology, in particular to a system for generating poetry in real time, which explores several aspects of our environment using data collected by IoT devices. It records various influencing factors such as temperature, humidity, light intensity, and noise level in relation to atmospheric conditions using environmental sensors, which then limit the poetic output to words / emotions / themes. Background of the invention:
[0002] Thanks to advances in AI, IoT, and analytics, artists can now apply new frameworks for creative implementations such as variation writing, environmental perception, and interactivity. These innovations have opened up the possibility of integrating real-time data into creative workflows, allowing art adapted to environmental variables like location and time to offer users an unprecedented experience. While this is progress, many solutions are still limited to text that is not dynamically generated, such as poetry that reacts to environmental conditions.
[0003] Legacy systems on which creative AI or generative art are based, on the other hand, typically rely on static datasets or pre-defined themes. This severely limits the timeliness and contextual awareness of the content these models generate, hindering the production of below-the-line content based on the user's location. To create a perfect dynamic poetry generation system that reads environmental data in real time, allowing you to prune the bushes before they grow into trees like the one above, we first need to learn how others have done it and what problems they've encountered.
[0004] In computational linguistics, this was the first implementation of dynamic text generation: techniques were developed that could automatically create sentences and even paragraphs based on input variables. These systems were highly rule-based and operated on the basis of syntactic structures or simpler lexicons without data from the environment. These systems produced text in one go and in a deterministic manner and were therefore insufficient as dynamic applications. Automatic sentence generation was already used in early educational tools, but these solutions were only capable of generating highly generic content with limited adaptability in the permissible interface, which was tailored to end-user input. Because these solutions could not handle the load of external sensory input, they were unable to provide output with valid context.Although they represented a valuable leap forward in machine-generated text production, the Poems for Robots model was not prepared to respond to data that changed in (often near) real time, which is necessary for developing a truly interactive poetic experience.
[0005] This achievement in AI, however, led to advances in text generation models based on sequence modeling: RNNs and Transformer networks were able to analyze massive datasets very effectively and learn how to sequentially generate further words that mimic human writing. These models make collaborative writing tools such as chatbots, content generation systems, and creative tools much more sophisticated. Until now, we essentially couldn't write to computers in human-like prose or poetry. While these systems represent a significant improvement over previous achievements, they still rely to some extent on input and linguistic patterns found in corporate language. They are still limited by a lack of context sensitivity.Newer AI text generators capable of writing emotionally charged sentences draw their content from static, outdated datasets and have been conditioned to be empathetic and able to adapt to environmental factors determined by other variables over time. In the world of creative writing, this has resulted in texts that are often detached from the user's current situation and inhibit real-time interactivity / conceptuality.
[0006] Another area where significant technological advances have been made is interactive installations that use IoT data as input for artistic output. However, they still have shortcomings in many respects. Some art installations incorporate sensors that react to the external environment. For example, they change visual displays or music depending on environmental influences; kinetic sculptures use motors and actuators controlled by computer programs, and so on. All of this creates immersive experiences in public spaces. For example, these installations contain sensors to measure environmental factors such as temperature, humidity, and light, and react by changing their color schemes or making other real-time adjustments using soundscapes or some form of artistic expression.In general, these systems are binary: they have some kind of input (usually visual or auditory) and output, but language itself is very rarely part of this poetic constellation. Furthermore, these solutions are often constructed in a closed environment tailored to the specific location and are not portable for more general applications or individual devices. Due to the limited reach of these installations, their adaptability to text-based interactive art, which generates corpus-calibrated linguistic outputs and interacts spontaneously with users via language, is restricted.
[0007] Few systems incorporate real-time data into the content generation cycle of literature or poetry. These systems are based either on predefined templates and forms of manual pre-programming, as there has been very little success in developing AI capable of producing literary constructs, even when a human writes a limited number of poems. The static variant includes strictly literary applications, such as "poetry bots," which randomly select or recombine lines of poetry without real-time interactivity or adaptation to the environment. This is particularly true where these systems rely on databases of hand-picked lines or phrases, which they then insert into pre-assigned forms based on themes deemed relevant.These bots, however, lack flexibility, as they can only generate content based on a pre-programmed argumentative structure, which is static compared to the dynamic creation of new content from external sources without constant manual intervention by humans. And finally, we need time: as contextualized poets and readers living in a landscape of real-time stimuli, freely malleable machines can never be them (or us), because they lack firsthand life experience; they are merely computers running complex AI models.
[0008] Efforts to integrate environmental data with text output have historically been limited and often experimental, such as experiments transforming smart homes into text assistants that prompt changes in pace or respond based on an established, long-term pattern. At best, these apps are informative and functional applications that objectify their information but do not create a comprehensive life experience. Environmental sensors for home assistants: As previously discussed, smart home assistants can detect some environmental signals, such as room temperature or light levels, but their output is mostly limited to basic updates and instructions like changing the thermostat, turning off the lights, etc.This deficit of creative linguistic interaction, coupled with a limited utility-oriented approach as opposed to numbing poetry emerging from real input, creates a gap between existing technology and situationally appropriate, emotionally evolving verses.
[0009] Furthermore, applications that offer near real-time text output often struggle with the quality and relevance of the material they produce. Environmental data is always situation- and context-dependent: a temperature increase means something different in an Arctic village at midnight than it does for a poor farmer on the worst day. These systems can easily misinterpret signals, even if we don't want to interpret them in the simplest way: "Here are two related data points → this system must have something to give you more accurate information." For a poetry-generating system, this lack of nuance in understanding environmental data can be very limiting, since poetry is about emotional connections and contextual relevance, which require a feel for the atmosphere.Therefore, current solutions struggle to create texts that are synchronized with both objective weather and subjective human perception, which significantly limits their capabilities in real-world applications.
[0010] To overcome these challenges, a system for real-time poetry generation using IoT environmental data must have a framework that can not only capture raw data but also interpret it in a way that reflects human experiences, particularly mood, emotions, and themes. However, existing models interpret environmental variables in isolation and are not designed to support creative, expressionistic translation into text. Furthermore, the techniques for linking various environmental factors (temperature, sound, and light intensity) with potential poetic themes / tones need significant further development in future work. Implementing these elements into the creative process will enable the development of an interactive and contextually appropriate poetry generator that responds in real time to what users see around them.This is a superior alternative to current solutions that use static data or only work in certain situations and have little linguistic versatility, but suggests a comprehensive artistic integration of IoT technology into literary applications. Summary of the invention:
[0011] The AI-based mechanism for generating poems to form the Dynamic Poetry Creation System using IoT environmental data consists of an environmental sensor network, a processing and analysis unit for sensor data received in real time from its modules, a custom-designed poetry writing engine powered by AI methods integrated through a proprietary LSTM model, and all other necessary specialized units. An environmental sensor collects real-time data at specific physical points of interest on campus (temperature and humidity sensors are one example, but sound levels and light intensity are also examples) and transmits it as raw logs of complex events to a docking unit. The final unit is an interpretation tool that reads the data to determine aspects such as mood, rhythm, and even theme.These are elements used as parameters in the poetry generation engine. Using a dynamic vocabulary and syntax database, the engine builds upon environmental inputs to both construct lines of poetry influenced by media commercialization and structure content accordingly, thus producing site-specific poetic compositions. These are displayed on an output interface, such as a digital screen or an audio speaker system, to convey existing environmental changes as immersive, real-time experiences that the poem attempts to evoke. By inputting environmental factors into the engine's dynamic vocabulary and syntax database, lines of poetry are generated in a way that reflects the commercialization of the media, and the content is structured accordingly, resulting in site-specific poetic compositions. The existing environmental change is represented, e.g.,through a connected output interface, a digital screen or an audio speaker system, which creates them as real-time experiences, tested through poems.
[0012] Another goal is to present poetry in an entertaining and accessible way, using visual or auditory representations. This includes incorporating the synchronization of external sensory elements, such as ambient lighting or sound adjustments, to create other forms of input that deliver multisensory experiences, amplifying the emotional impact of poetry. Furthermore, the new poetry application aims to provide users with personalized configurations that allow them to save, share, or even store poems created under specific environmental conditions, so that the generated artistic creation forms a lasting and intimate connection between them.
[0013] Ultimately, it aims for dissemination in disciplines beyond functional automation and into the realm of artistic creation, which has not previously been open to application in an interactive, real-time experience, as proposed by truefuhlen, by introducing an innovative IoT (Internet of Objects) approach that disrupts traditional poetry. Inspired by the theory of dynamic sculptures, this invention hopes to engage users in their environment in more creative ways and, consequently, promote environmental awareness by reimagining, through literary art, the way people not only perceive but also interact with their surroundings. BRIEF DESCRIPTION OF THE FIGURE
[0014] These and other features, aspects, and advantages of the present invention will be better understood if the following detailed description is read with reference to the accompanying drawing, in which the same symbols consistently represent the same parts. The following applies: Fig. Figure 1 shows a block diagram of a dynamic system for poetry creation based on IoT environmental data.
[0015] In Fig.Figure 100 is a block diagram of a dynamic system for generating poems based on IoT environmental data. The system comprises: an environmental sensor module 102 with a variety of sensors configured to capture real-time environmental data, including at least temperature, humidity, sound intensity, and light intensity, with each sensor outputting specific data signals related to the surrounding environmental conditions; a data processing unit 104 communicating with the environmental sensor module, the data processing unit configured to receive, aggregate, and analyze data signals from the variety of sensors to derive context-specific parameters related to mood, theme, and rhythm for poem creation, with the parameters being continuously updated based on variations in the real-time environmental data.a parameterization machine 106 embedded in the data processing unit, wherein the parameterization machine is configured to convert the context-specific parameters into a predefined set of dynamic variables associated with thematic constructs, emotional tones and linguistic styles, wherein the parameterization machine adapts the dynamic variables in response to detected changes in environmental data;a poem generation machine 108 communicatively connected to the data processing unit, wherein the poem generation machine comprises a vocabulary database 108(a) organized by themes and tones and a syntactic arrangement module configured to select certain vocabulary elements from the vocabulary database and arrange them according to the dynamic variables received from the parameterization machine, wherein the syntactic arrangement module generates structured poem lines in real time based on the environmental conditions;and an output interface 110, which is operationally connected to the poem generation machine, wherein the output interface is configured to display or audibly present the generated poem lines to a user, the output interface continuously updating the displayed or recited poems synchronously with changes in the environmental data. The present invention comprises artificial intelligence and machine learning models for rapid self-learning based on historical data and the environmental data.
[0016] In one embodiment, the environmental sensor module 102 also includes an air quality sensor configured to detect pollutant concentrations in the air, and wherein the data processing unit integrates pollutant data as an additional parameter to influence the heaviness or lightness of the poetic tone and to adjust the mood according to the air quality levels.
[0017] In one embodiment, the data processing unit 104 includes a machine learning-based interpretation module that refines the thematic and emotional parameters over time based on historical environmental data patterns, so that the system adapts to certain environmental trends and adjusts the poem themes accordingly.
[0018] In one embodiment, the parameterization engine 106 is also configured to assign weighting values to each environment parameter based on the degree of change detected in real-time data, whereby the weighting values influence the dominance of certain themes or tones in the generated poetry, thus enabling an improved response to rapid or subtle environmental changes.
[0019] In one embodiment, the output interface 110 includes an option for storing generated poetic compositions in a user-accessible archive, so that users can save poems created under certain environmental conditions for later reference, with each stored entry being labelled with the associated environment data parameters.
[0020] In one embodiment, the environmental sensor module 102 also includes a geolocation unit for providing real-time location data, wherein the geolocation data influences the vocabulary selection within the lyric generation engine to reflect location-based environmental contexts and thereby enable location-specific lyric generation.
[0021] In one embodiment, the data processing unit 104 also includes an alarm-based adaptation subunit that triggers a theme change when environmental parameters exceed predefined thresholds, wherein the theme change alters the vocabulary and syntactic arrangement in the generated poetry to reflect extreme environmental changes such as adverse weather conditions or increased sound intensity levels.
[0022] A structural enclosure, which necessarily includes a networked hardware infrastructure, is used to dynamically generate poetry based on IoT environmental data. These components consist of an environmental sensor array, a data processing unit, a parameterization engine or component module processor, a vocabulary management circuit and syntax assembly, and a device output display interface. The environmental sensor array has sensors for streaming live data from up to 10 different environmental sources: temperature, humidity, sound, and light. Each enclosure contains an embedded sensor, and each sensor is calibrated to detect variations in its environmental conditions and transmit data signals to a processing unit via an internal signal bus.This data processing unit includes a central processing hardware component to aggregate sensor outputs and store them as theme factors that reflect the effects of environmental inputs. The data processing unit comprises a storage module and a logic-based data filter for detecting environmental changes associated with key parameters that define emotional states, such as mood-shifting theme elements, which are transmitted to the parameterization engine in real time.
[0023] In this system, the parameterization engine features configurable logic blocks that evaluate environmental data streams and assign weights to these factors based on detected fluctuations, affecting poetic themes. A control bus allows these logic blocks to communicate and apply the effects of a changing data stream to theme selection, enabling dynamic vocabulary adjustments. Structured Vocabulary Memory Array – This memory module is essentially an array of all units stored on a chip with structured vocabulary memory; words are categorized by tone and intensity. This indexed, fast-access vocabulary array is connected to the syntactic arrangement processor via a high-speed access bus, allowing for the retrieval and assembly of theme-appropriate words during poem generation.The syntactic order processor, an array of microcontrollers, orders vocabulary elements based on the properties observed in structured lines of poetry using a compiled syntax library; it updates itself to adapt its own forms to conspicuous environmental conditions.
[0024] Regarding output, the display interface consists of a high-resolution screen embedded in the housing, connected to an image processor that converts text data into human-readable formats to display generated poems. There is also a user-customizable display where poem creations can be stored under specific environmental conditions. An air quality sensor is another element of the environmental data array, with an analog-to-digital conversion module that provides an assessment of mood-damaging pollution and, based on this, allows for variation in the poetic tone. The housing also contains an audio output unit with a speaker array and audio amplification circuitry, enabling the system to recite poems aloud and add a multi-sensory dimension.Additionally, an external communication port of the housing is connected to a wireless module to retrieve location-specific data and transmit it to remote user devices after poem processing.
[0025] The solution is an IoT-based system for collecting and processing environmental data to generate dynamic figures. The system primarily receives input from an environmental sensor module capable of measuring temperature, humidity, sound intensity, brightness, and air quality. These sensors are wirelessly connected to a central data aggregation unit via peer-to-peer communication protocols (Wi-Fi or Bluetooth), depending on which performs best within the network limitations. All sensors transmit their data to the central data aggregation unit, which then forwards it to the data processing unit every 10 seconds.
[0026] The data processing unit processes the transmitted information to determine the environmental conditions and outputs this information to adjust subsequent phases during the poem's creation. Warmth and vibrancy might be associated with higher maximum temperatures, while solitude or introspection might occur at minimum temperature levels. Similarly, a high noise level might evoke a sense of urban bustle or chaos, while participants speak in a cheerful, contemplative tone at quieter noise levels. This processing unit includes a parameterization engine that measures each environmental factor as absolute numbers on predefined thematic scales.
[0027] The specified parameters are then passed to a poetry generation engine, which includes a vocabulary database and a syntactic arrangement module. The vocabulary database is a diverse lexicon categorized by themes, emotions, and sensory perceptions. It selects words from specific word categories based on the input parameters, which are conditionally usable in real time under certain environmental conditions. The syntactic arrangement assembles the selected language into formatted lines of poetry, using pre-made templates for a comprehensive repertoire that includes haikus, sonnets, and free verse. This module ensures that the poem conforms to grammatically and aesthetically correct structures.
[0028] The generated poetic content is then output to a display interface. This interface can be a digital screen or a speaker system that displays / reads the poems aloud, allowing people to hear and experience poetry in their environment. Alternatively, the output can be shared via a mobile app, enabling users to archive or share poems created at specific times in their vicinity. The display configuration is enhanced by options for connecting and synchronizing additional IoT devices, such as color-changing LEDs that match the mood of the generated poem; in other words, a comprehensive sensory poetry experience.
[0029] The Dynamic Poetry Creation System consists of a housing containing the environmental sensor module, the data processing unit, and the poetry generation engine, which is integrated with a display interface. The housing protects the sensitive electronics from environmental factors, enabling effective operation in a variety of environments: public spaces, parks, and indoor installations. A sensor module on the top of the housing optimally captures and records environmental data, while the main compartment contains all the data processing units and the poetry generation engine. Users can read these poems on the display interface, which shows them as fluid text in real time via analog outputs located on the front of the structure.The embedded connectivity module also enables connections with external sensors for additional functions such as lighting systems or smartphone applications, thus improving the interactivity of the experience.
[0030] The present invention offers a unique way to combine IoT technology and creative expression: it generates poetic creations influenced by real-time environmental contexts. The current implementation of AmbiVerse serves as a novel tool that helps readers interact dynamically with poetic texts, opening up new possibilities for education, public art installations, and customizable literary worlds. REFERENCES 100 A dynamic system for creating poems based on IoT environmental data. 102 Environmental sensor module 104 Data processing unit 106 Parameterization Engine 108 Poetry Generation Machine 108a Vocabulary database 110 Output interface
Claims
[1] A dynamic system for creating poems based on IoT environmental data, consisting of: an environmental sensor module having a plurality of sensors configured to collect real-time environmental data, including at least temperature, humidity, sound intensity, and light intensity, each sensor outputting specific data signals related to the environmental conditions in the surrounding area; a data processing unit in communication with the environmental sensor module, the data processing unit configured to receive, aggregate, and analyze data signals from the plurality of sensors to derive context-specific parameters associated with mood, theme, and rhythm for poem creation, the parameters being continuously updated based on variations in real-time environmental data; a parameterization engine embedded in the data processing unit, the parameterization engine configured to convert the context-specific parameters into a predefined set of dynamic variables associated with thematic constructs, emotional tones, and linguistic styles, the parameterization engine adapting the dynamic variables in response to detected changes in the environmental data; a poetry generation engine communicatively connected to the data processing unit, the poetry generation engine comprising a vocabulary database organized by themes and tones and a syntactic arrangement module configured to select specific vocabulary items from the vocabulary database and arrange them according to the dynamic variables received from the parameterization engine, the syntactic arrangement module generating structured poetic lines in real time based on the environmental conditions; and an output interface operatively connected to the poem generation engine, the output interface configured to display or audibly recite the generated lines of poetry to a user, the output interface continuously updating the displayed or recited poems in synchronization with changes in the environmental data. [2] The dynamic poetry creation system according to claim 1, wherein the environmental sensor module further comprises an air quality sensor configured to detect pollutant concentrations in the air, and wherein the data processing unit integrates pollutant data as an additional parameter to influence the heaviness or lightness of the poetic tone and adjust the mood according to the air quality levels. [3] The dynamic poem generation system according to claim 1, wherein the data processing unit comprises a machine learning-based interpretation module that refines the thematic and emotional parameters over time based on historical environmental data patterns, so that the system adapts to certain environmental trends and adjusts the poem themes accordingly. [4] The dynamic poem generation system of claim 1, wherein the environmental sensor module further comprises a geolocation unit for providing real-time location data, the geolocation data influencing vocabulary selection within the poem generation engine to reflect location-based environmental contexts. [5] The dynamic poem generation system of claim 1, wherein the data processing unit further comprises an alert-based adaptation subunit that triggers a theme change when environmental parameters exceed predefined thresholds, the theme change altering the vocabulary and syntactical arrangement in the generated poems to reflect extreme environmental changes, such as adverse weather conditions or increased sound intensity levels.