AI Space Utilization Prediction Using Multivariate Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques fail to model or measure space utilization at a granular level, considering variables such as day of the week, time of day, design aesthetics, occupation, team or department association, and user duties, leading to incomplete understanding of space usage.
Innovation Solution
A multivariate model using artificial intelligence techniques, specifically a Mixture of Experts Neural Network (MENN), is trained with location data from mobile devices and space metadata to predict space utilization, accounting for various variables and providing granular insights into space usage patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing techniques are used to measure space utilization, then the measurement process is simple, but the measurement precision and granularity are insufficient
Solution Approach 1:
The patent segments space utilization measurement into multiple dimensions including temporal variables (day of week, time of day), spatial variables (specific locations within space), and contextual variables (design aesthetics, occupation, team association). This segmentation enables granular measurement by breaking down the overall utilization metric into component parts that can be analyzed independently and combined to provide comprehensive insights.
Solution Approach 2:
The patent introduces multiple additional dimensions for measuring space utilization beyond simple occupancy counts. These dimensions include time-based dimensions (different days and times), spatial dimensions (specific areas within the space), and attribute dimensions (design features, user characteristics). By adding these dimensions, the system achieves much higher measurement granularity.
2Measurement precision
If a multivariate model with multiple variables is used to predict space utilization, then the prediction accuracy and insight granularity are improved, but the model complexity increases
Solution Approach 1:
The multivariate model is segmented into multiple independent variable components (temporal variables, spatial variables, design variables, user variables) that can be processed and analyzed separately before being combined. This segmentation makes the complex model more manageable by allowing each variable type to be handled with appropriate processing techniques.
Solution Approach 2:
The patent uses an electronic processor as an intermediary to handle the complexity of the multivariate model. The processor receives data from multiple sources, processes the various variables through the MENN framework, and produces synthesized predictions. This intermediary handles the computational complexity, allowing the model to incorporate many variables without making the system itself more complex.
3Quantity of substance
If location data from multiple mobile devices is collected and processed, then the data completeness and model training quality are improved, but the data processing complexity and computational requirements increase
Solution Approach 1:
The electronic processor is designed with multi-functionality to handle various aspects of data processing including data collection from multiple devices, data validation, feature extraction, model training, and prediction generation. This universal processor handles all data processing tasks through a unified framework, reducing the need for separate specialized systems for each processing step.
Solution Approach 2:
The system implements self-service through automated data processing pipelines where the electronic processor automatically collects location data, processes it through the MENN framework, trains the model, and generates predictions without requiring manual intervention. The system serves itself by automatically managing the entire data processing workflow from raw data to actionable insights.
Data Source
AI summary
Methods and systems for training a multivariate model predicting utilization of a space. One method includes receiving, over a period of time, signals from each of a plurality of mobile devices located in the space and generating, based on the signals received from each of the plurality of mobile devices, a plurality of location data points for each of the plurality of mobile devices, each of the plurality of location data points for a mobile device including a timestamp and a position within the space of the mobile device. The method also includes accessing metadata of the space, and using, with an electronic processor, the plurality location data points for each of the plurality of mobile devices and the metadata of the space to train machine learning engine. In addition, the method includes predicting a utilization of the space using the machine learning engine.