AdHocProx: Sensing Mobile, Ad-Hoc Collaborative Device Formations using Dual Ultra-Wideband Radios

Document Title

AdHocProx: Sensing Mobile, Ad-Hoc Collaborative Device Formations using Dual Ultra-Wideband Radios

Document Information

• Topic Area: Device interaction technology, multi-device collaboration, sensing technology
• Keywords: Multi-device collaboration, ultra-wideband sensing, proxemics, inside-out tracking, device interaction, gesture recognition

Research Background and Problem

• Identified Problems or Challenges:

  • Configuring multi-device collaboration often requires users to perform cumbersome operations, such as relying on WiFi, Bluetooth, or cloud services, which disrupts user workflows and natural social interactions.
  • Existing cross-device interaction technologies lack automatic discovery and configuration of relative positions and orientations between devices, limiting dynamic and ad-hoc collaboration scenarios.
  • The design of collaboration between mobile devices lacks a dynamic understanding of "proxemics" and "social distances."

• Significance:

  • With the widespread use of mobile devices, achieving fast and simple multi-device collaboration is crucial for enhancing work efficiency and improving user experience.
  • Enabling such collaboration in environments outside of infrastructure-heavy "smart rooms" will expand the applicability of collaboration technologies.

• Research Motivation and Related Work:

  • Inspired by sociological theories (Proxemics, Micro-Mobility), this study aims to optimize the device collaboration experience from the perspective of social interactions and user behavior.
  • Technologically, ultra-wideband (UWB) sensing and inside-out device tracking technologies are maturing, but their potential in dynamic device collaboration remains underexplored.

Solution

• Proposed Solution:

  • The AdHocProx system uses devices equipped with dual UWB radios to achieve "inside-out" sensing of relative distances and orientations between devices.
  • By leveraging UWB Time-of-Flight measurements, along with devices equipped with capacitive grip sensors and inertial measurement units (IMUs), the system can detect relative positions, orientations, and user gestures or grip states.

• Innovations:

  • Eliminates traditional configuration steps (e.g., WiFi pairing) in multi-device collaboration, enabling devices to automatically detect and configure their relative positions.
  • Utilizes dual UWB radios for sensing relative orientation, a feature not yet widely applied in commercial devices.
  • The system requires no external fixed beacons or coordinating devices, making it suitable for dynamic and ad-hoc collaboration scenarios.

• Implementation Steps and Key Technologies:

  • Hardware Design: Utilizes two ESP32 microcontrollers, two DW3000 UWB radio modules, and four capacitive touch sensors.
  • Data Processing: The system collects UWB distance measurement data, gesture data, and IMU signals, combining them with capacitive sensor calibration through machine learning (random forest classifier) to identify device orientations.
  • User Interface: A "Portal System" was designed to display and manage shared content between devices, providing visual feedback based on user interaction behaviors.

Research Outcomes

• Specific Outcomes:

  • The AdHocProx system can identify the arrangement and orientation of dynamic device arrays with 95% accuracy.
  • Designed and implemented four interaction techniques (Move, Copy, Pan, Note) to support position-aware collaboration between devices.
  • Established foundational theoretical support for observing device collaboration behaviors and collected a sensor signal dataset for system evaluation.

• Advantages:

  • Compared to traditional multi-device interaction systems, AdHocProx offers higher dynamic adaptability and lower deployment costs.
  • Does not require external network connections or beacons, making it suitable for more scenarios (e.g., ad-hoc team collaboration or non-fixed lab environments).
  • Gesture-based interaction is more natural, combining sociological theories and user behavior insights.

• Experiment or Evaluation Results:

  • User feedback revealed that interaction designs based on device tilt and relative positions (e.g., the Note function) enhance collaboration efficiency.
  • Offline evaluation showed that the capacitive sensor calibration technique resolved UWB signal errors caused by grip, significantly improving classifier accuracy.

• Limitations and Future Directions:

  • Currently limited by UWB radio bandwidth and battery consumption issues; hardware improvements will be a key focus in the future.
  • Further research is needed on devices acting as user proxies and addressing privacy and security challenges.
  • Future work could extend to remote collaboration or integrate the system with wearable device-based environmental sensing.
  • Explore more interaction techniques based on automatic sensing of device arrays, including hybrid collaboration scenarios or dynamic content delivery based on user contexts.