Swipe-to-Locate: Enhancing Real-World Object and Event Localization

Unmet need
Conventional techniques for mapping the location of real-world objects or events are successful at identifying the location of static objects at a gross level. Mobile devices such as smartphones typically include GPS for navigation and displaying a device’s position on a map. However, localization of small or time-varying events has yielded poor accuracy. Accordingly, there is a need for improved systems that quickly and efficiently define the geographic position of events or objects near the user.

Technology
Researchers at Duke have developed a mobile application that enhances the localization of real-world objects, events, and behaviors through user interaction and mobile sensing. The system determines the geographic position and orientation of a device, allowing users to define the location of an object or event with a simple swipe on the touchscreen. By tracking the direction and distance indicated by the user’s gesture in relation to the device’s position, the application provides accurate localization even for events that users cannot approach closely. This method aggregates data from multiple users, enhancing localization accuracy through crowdsourcing. The application offers functionality such as defining an object's geographic position based on satellite positioning, the device's sensors, or wireless signals. The app can further gather user-specified details (e.g., object name) and communicate these with other devices, facilitating broader data sharing and event tracking.

Other Applications
The app has wide potential applications for participatory reporting of events, such as infrastructure issues. It can also be used by institutions to evaluate the effectiveness of public policies or regulations, such as event control or safety monitoring.

Advantages

• Low-cost and near-continuous solution for event, behavior, and object localization
• Gesture-based input enables localization without needing to be near or directly face the target
• Enhanced localization accuracy by combining user gestures and device orientation
• Easily scalable with the potential to integrate crowdsourced data from multiple users
• Outperforms traditional mapping systems in localization precision and coverage