Indoor/ Outdoor Pedestrian Navigation with an Embedded GPS/ RFID/ - - PowerPoint PPT Presentation

Indoor/ Outdoor Pedestrian Navigation with an Embedded GPS/ RFID/ Self- contained Sensor System Masakatsu Kourogi, Nobuchika Sakata, Takashi Okuma and Takeshi Kurata National Institute of Advanced Industrial Science and Technology (AIST)

Research background (1)

Location/ direction-based Web services are widely available to provide local information such as maps, weather and nearby

• transportation. (ex. Google Maps and

Yahoo! Maps) Portable PCs and PDAs (smart phones) are capable of rendering 3D urban landscape. (ex. Google Earth and Pocket Cortona)

Research background (2)

Combination of location/ direction- based services and a suite of 3D mapping software will provide highly intelligent navigation system. It is essentially important to acquire accurate position and direction to enable such navigation system.

Research targets

In both indoor/ outdoor environments, to achieve stride-level accuracy of positioning method.

It is realized by dead-reckoning method combined with RFID and GPS.

To be implemented by an embedded computing system and provide pedestrian navigation services.

Proposed method:

Dead-reckoning based on walking locomotion

Self-contained sensors (gyroscope, magnetometers and accelerometers) realize dead-reckoning based on human walking locomotion.

Partially proposed by our previous researches. Dead-reckoning will work in both indoor/ outdoor environments.

Proposed method:

Dead-reckoning based on walking locomotion

Accuracy of dead-reckoning is vulnerable to accumulation of step- wise error and thus requires external sources of information about absolute position to correct such error.

First, error model of dead-reckoning is required. Second, GPS and active RFID tag system are used as external position correction.

Proposed method: Error model of dead-reckoning

The true azimuth The estimated azimuth The true stride The estimated stride error The previous position

The error of dead-reckoning is composition of that of azimuth and of stride.

Proposed method: Error model of dead-reckoning

• Combination of error caused by azimuth and stride are

approximated by Gaussian distribution.

approximation 2-D Gaussian distribution

Distribution of the previous position Distribution for azimuth Distribution for stride

i

l

i

θ

1-step of dead-reckoning

Known to be Gaussian Known to be Gaussian

Next position

Proposed method: Error model of dead-reckoning

Kalman filter is used to update the estimation of position and velocity.

• The state vector:
• Update equations:

[ ]

T y x t t t

t t

v v y x = s

) (

| 1 t t t t t t

s O K s s − + =

+ 1

) (

−

+ =

t t t t

R P P K

t t t t t

P K P P − =

+ | 1

position velocity Estimated from acceleration during the walking locomotion

t

K

Kalman gain

t

P

Covariance matrix

• f the state vector

t

R

Covariance matrix of the error of observation

t

O

Observation of the state

Proposed method: Estimation of pedestrian velocity

1 2 3 4 5 6 7 8 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Forward acceleration gap [G] Speed [Km/h]

Subject A Subject B Fitting line (A) Fitting line (B)

• The velocity and acceleration gap (amplitude) are

highly correlated and estimated from the other.

Proposed method: Combination with external sources (1) GPS is used in outdoor environment.

Error of GPS has three components:

Error caused by multipath effect Offset error caused by signal delays. Random error (represented by Gaussian)

Multipath error can be reduced with combination of dead-reckoning.

GPS data out of 95% area is excluded for computation GPS data included for computation The previous position by dead-reckoning 95% confidence area Estimated by dead-reckoning

Proposed method: Combination with external sources (2) GPS is used in outdoor environment.

Offset error can be measured by the fixed observation station whose location is exactly known. Thus, remaining random error can be handled within Kalman filter framework since it has normal distribution. Two measurements by GPS hints the position and velocity in the state vector.

Proposed method: Combination with external sources (3)

Active RFID tag system is used to correct the pedestrian position.

• Error of position can be represented by the

Gaussian distribution.

Floor A RFID tag reader Range of position

0.6m 1.5m 1.3m The height of user’s waist where the RFID tag is attached Reachable range of the ID signal Connected via wireless network

Implementation: Outlook of the system

The total system is implemented in an embedded computing system.

• A browsing system is

separately implemented.

• HMD system and handheld

system are implemented.

Embedded Computer Self-contained sensors Handheld PC GPS Orientation Tracker RFID Tag Camera Camera

HMD user HMD user Handheld display user Handheld display user

HMD Orientation Tracker

Implementation: Diagram of the system

Dead-reckoning GPS positioning Self-contained sensors

GPS module

Embedded computing system

Positioning engine

Estimation results of position and direction (in NMEA0183 or CSV format) Attitude tracker Data distribution engine Google Earth (Navigation application) Browsing system (SONY VAIO type U) CCD camera JPEG image Equipped with user’s hip

RFID tag

Equipped with browsing PC (via Wireless network) RFID positioning From the RFID tag system From/To the control server Position/direction

• f other users

Adjustment request (via Wireless network)

Implementation: Diagram of the system

Google Earth CGI code JavaScript (Asynchronous)

Navigation browsing system

Dynamically generated KML data

Location/direction Web server Remote Web server (with SQL database) CGI code Query with location/direction Nearby contents CGI code

Query Location/ direction

Embedded system

Demo video: Our system in actions (Indoors)

Demo video: Our system in actions (Outdoors)

Experiments: The ground truth vs estimation

Setting for experiment:

368.1 meter route (outdoor: 247.2 meter, indoor: 120.9 meter) Two RFID readers are placed in the building to correct user’s position. Five subjects traveled along with the same route. Estimations by the proposed method were compared to the ground truth of the route.

GPS only DR only Our method Ground truth Start here Due to drift errors in gyrosensors, the result of dead-reckoning deviates. Differentiation of GPS results adjusted the error in direction. 25m GPS only O u r m e t h

• d

DR only Ground truth

Experiments: The ground truth vs estimation

Error graph along with walking distance

5 10 15 20 25 30

50 100 150 200 250 300 350 400 Walking distance (m) Averaged error (m) GPS only DR only Our method

Outdoor Indoor RFID adjustments GPS results are used to adjust errors in direction/position GPS results are discarded Elevator detection adjustment

Practical usage of the system: Openhouse situation with kids

23 kids have experienced our prototype system.

• No prior training or

calibration is required.

• User’s height is only

parameter required by the system.

• The system worked

well even if kids moved in unexpected manners.

Conclusion

GPS/ RFID/ Dead-reckoning integrated positioning method is proposed. Embedded pedestrian navigation system is implemented with the proposed method. Accuracy of the proposed method is shown to be 5-10% of total walking distance. Kids can play with the system in the

• pen house situation.