SLIDE 1
LookUp Enabling Pedestrian Safety Services via Shoe Sensing - - PowerPoint PPT Presentation
LookUp Enabling Pedestrian Safety Services via Shoe Sensing - - PowerPoint PPT Presentation
LookUp Enabling Pedestrian Safety Services via Shoe Sensing Shubham Jain In the last decade, more than 47,000 people died while Pedestrians account for 15% of all traffic fatalities! walking on American streets! Existing Awareness Cues TIME
SLIDE 2
SLIDE 3
Pedestrians account for 15% of all traffic fatalities! In the last decade, more than 47,000 people died while walking on American streets!
SLIDE 4
WalkSafe: A pedestrian safety app for mobile phone users who walk and talk while crossing roads (Wang et al) The Benefits of Dense Stereo for Pedestrian Detection (Keller et al) TIME Magazine, February ‘14
Existing Awareness Cues Existing technology-based pedestrian safety solutions
New York City
SLIDE 5
Detects when a pedestrian is entering the roadway What if your phone could sense when you are entering the street, and warn you if you are distracted..
SLIDE 6
Sensing Ground Features that Separate Street and Sidewalk
SLIDE 7
GPS accuracy is not enough for street-sidewalk distinction!
~35 meters Actual Path Walked GPS Trace
SLIDE 8
Shoe-mounted Inertial Sensing
Accelerometer Gyroscope Magnetometer Bluetooth
SLIDE 9
Ground Profiling via Shoe Mounted Inertial Sensors
SLIDE 10
Measuring changes in foot inclination
Gx (0g) Gz (-1g) G [0g, 0g, -1g] Gx (0.1g) Gz (-0.9g)
10°
G [0.1g, 0g, -0.9g] Gx (0.3g) Gz (-0.7g) 30°
G [0.3g, 0g, -0.7g] Accelerometer based Gyroscope based
Angular Velocity
Angle (Pitch)
SLIDE 11
Low Pass Filter High Pass Filter
Pitch (α)
Complementary Filter
Accelerometer Gyroscope
Complementary Filter
Calculating pitch by combining accelerometer and gyroscope measurements
SLIDE 12
Phases Of A Walking Cycle
SLIDE 13
Transition Patterns
SLIDE 14
Sidewalk - Street Transition via a Curb
SLIDE 15
Evaluating our system in the real world
SLIDE 16
Manhattan Visual Rundown
Typical Ramp Typical Curb Crowded Environment Sidewalk obstacles
SLIDE 17
SLIDE 18
Number of participants: 22 Test locations: New York City and Turin Total distance covered: 112.5 miles Number of crossings: 1670 Total walking duration: ~ 80 hours
Dataset Statistics
SLIDE 19
Detection Latency
SLIDE 20
Ground Truth Window
True Positive: Detection that occurs inside the window False Positive: Detection that occurs
- utside the window
SLIDE 21
Locations for Detections
SLIDE 22
System Performance
SLIDE 23
Conclusion
We have devised a sensing technique that profiles ground gradient via shoe-mounted inertial sensors. We use these ground profiles to detect events when a pedestrian transitions from sidewalk to street via a ramp or curb. The transition detection approach relies on consistent sidewalk designs. It does not work well in suburban environments where the presence of sidewalks is not consistent. We observed that the performance could be sensitive to sensor mounting and can be improved with robust mounting designs. The proposed approach also relies on the instrumentation and power in shoes. In the future, we intend to build a reliable pedestrian to driver communication scheme.
SLIDE 24
… Thank You!
SLIDE 25