Predictive Location-Based Information Delivery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current location-based services only provide information based on a device's current location, limiting the relevance of the information sent, as they do not consider the device's predicted future locations.
Innovation Solution
A server collects location histories from user devices to predict future locations and sends information based on these predictions, allowing for actions to be taken when a device is likely to be at specific locations within a certain time frame with a specified likelihood, such as sending promotions or advertisements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If location-based services only use current location to send information, then the system complexity is low, but the relevance of information sent is limited
Solution Approach 1:
The system performs preliminary actions by collecting location history data in advance and building prediction models before actual information delivery is needed. This allows the system to predict future locations and send relevant information proactively, resolving the contradiction between information relevance and system complexity by preparing predictive capabilities beforehand rather than computing them in real-time
Solution Approach 2:
The system transitions from considering only the current location (one dimension) to incorporating historical location data and temporal patterns (multiple dimensions). By adding the time dimension and historical context, the system creates a multi-dimensional view of device behavior that enables much more relevant information delivery without prohibitively increasing complexity
2Reliability
If the system collects and analyzes location history to predict future locations, then the relevance of information is improved, but the data processing requirements and system complexity increase
Solution Approach 1:
The system applies partial action by focusing prediction efforts on specific high-value scenarios and locations rather than attempting to predict all possible future states. The prediction model concentrates computational resources on determining whether a device is likely to visit predetermined locations of interest, rather than exhaustively analyzing all potential future trajectories, thus improving reliability without excessive complexity
Solution Approach 2:
The system employs self-service mechanisms where the prediction model learns from historical data and automatically improves its accuracy over time without requiring manual intervention or complex reconfiguration. The model serves itself by continuously refining its predictions based on accumulated location history, reducing the operational complexity while maintaining high reliability
3Productivity
If the system sends information based on predicted future locations, then the effectiveness of location-based services is improved, but the computational resources and time required increase
Solution Approach 1:
The system performs preliminary computation by pre-building prediction models and pre-identifying likely future locations during off-peak periods. This allows the system to make rapid predictions when needed for information delivery, resolving the contradiction between service effectiveness and computation time by doing the heavy lifting beforehand
Solution Approach 2:
The system uses periodic action by updating prediction models at scheduled intervals rather than continuously re-computing predictions. This periodic approach maintains high service effectiveness through accurate predictions while reducing computational overhead by avoiding constant recalculation, thus balancing productivity with time efficiency
Data Source
AI summary
A server sends information to user devices based on not only the current location of those devices but also predicted future locations of those devices. A number of actions are associated with one or more predetermined locations, a predetermined maximum amount of time, and a predetermined minimum likelihood. When the server determines that a given user device is likely to be in one of the predetermined locations within the predetermined maximum amount of time with at least the predetermined minimum likelihood, the server performs the associated acts with respect to the user device. An example of such an action is sending a promotion or advertisement to the user device, e.g., as an SMS message.