SmartKG Logo

iBadge

Concept

 :: Home :: Concept :: Hardware :: Software :: Publications :: 


The iBadge is one of the crucial components in the Smart Kindergarten project. It is equipped with sensors such as microphone, localization sensor and temperature sensor. Kids and the teachers are supposed to wear the iBadge. The iBadge is not only a collection of sensors; it possesses enough computational power to process the sensed data. Two processors, Atmel ATMEGA and TI DSP C5416, share the workload according to their computational ability. The DSP processes the computational more expensive speech data and the magnetic sensor data, whereas ATMEGA is interfaced to sensors with simpler data structure and the Bluetooth module. The Bluetooth module provides the wireless connection to the rest of the infrastructure, to which all the collected and processed data is sent.

Some of the most critical design constraints of the iBadge include form factor, energy consumption, cost, design flexibility, and sensing capabilities. It must also be light and small enough for children to wear. A typical operational scenario for the Smart Kindergarten requires iBadges to operate for 4 to 6 hours a day.

The iBadge is partitioned in functional units (see also block diagram):

  • Localization Unit (LU): It is responsible for obtaining the precise spatial location of an iBadge based on an atomic multi-lateration process. 

  • Environment Sensing Unit (ESU): It delivers measurements about temperature, light intensity, humidity and pressure of the environment.

  • Orientation and Tilt Sensing Unit (OTSU): It includes two dual-axis accelerometers and a single tri-axis magnetometer, detecting motion and the presence of magnetic fields. The magnetometer measures the Earth’s magnetic field along three axes to determine the iBadge’s orientation. Together, these sensors can act as a tilt-compensated compass.

  • Speech Processing Unit (SPU): The SPU consists of a DSP, a codec chip, a microphone, and a loud­speaker. The iBadge can send either compressed speech (using G.722 wideband coding) or speech feature vectors extracted by iBadge doing the front-end processing. The speech feature vectors can directly be applied to the automatic speech recognition process.

  • Power Management / Tracking Unit (PMTU): The PMTU has the ability to deactivate functional units in order to conserve energy. Moreover, it monitors iBadge’s energy consumption. Using the PMTU, the iBadge can implement power-aware algorithms and battery-aware power management schemes to prolong battery life.

  • Wireless Communication Unit (WCU): The iBadge communicates over Bluetooth with the rest of the infrastructure. All collected and processed data is sent over this wireless interface.

  • Power Supply Unit (PSU): It provides three regulated voltages: 1.6V, 3.3V and 5V.

The PMTU provides iBadge with the ability to switch on/off different parts of the circuit to conserve power. On the ground planeside, the battery monitors acquire the currents drawn by the entire circuit and also separately keep track of current drawn from the Speech Processing Unit, the Sensing Units (Localization, Orientation/Tilt Sensing, Environment Sensing) and the Wireless Communication Unit. The iBadge uses shunt resistors to measure the current ranging from 0.1 Ohms to 0.39 Ohms with 1% tolerance, resulting in resolutions of 2.44mA/LSB to 0.63mA/LSB. Applying these measurements, iBadge is able to implement power aware algorithms or battery state aware power management schemes to prolong battery lifetime. The power concept figure depicts the hardware arrangement of PMTU. Because of the partitioning of the circuit, separate power and ground nets are deployed for each functional unit.


In the near future, we will investigate the optimal parameters for power management on the iBadge. This mainly includes finding the best ratio between on-badge computation and wireless data transmission, constrained as they are by a limited battery. 


This material is based upon work supported by the National Science Foundation under Grant No. ANI-0085773. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).
 

© 2002 by Networked & Embedded Systems LaboratoryUniversity of California, Los Angeles
Author: Ivo Locher