Wireless Body Sensor Networks

Recognizing human activities using Wireless Body Sensor Networks (WBSNs) has received considerable attention from both academic researchers and industry practitioners in the past few years. This task is particularly challenging because human activities are performed in not only a simple (i.e., sequential), but also a complex (i.e., interleaved and concurrent) manner in real life, and often involved in multiple users.

At the University of Southern Denmark, there are several research projects, led by Dr. Gu, investigating the fundamental problems of recognizing real-world, complex activities for both single and multiple users in WBSNs. They have designed a multimodal WBSN to capture observations for multiple users. It consists of a number of Imote2s, two RFID reader Motes, and an acoustic sensor node as shown in the figure above. An Imote2 node captures movement, environmental temperature, humidity and light. An RFID reader Mote captures object use within several centimeters, and it consists of a MICA2DOT Mote and a coin-size short-range RFID reader. It also uses an audio recorder as an acoustic sensor to capture sound.

They then conducted data collection by two subjects in a home environment. Each subject wore a set of wearable sensors and performed various activities by following his/her own step for each activity. Some snapshots are shown in the figures below.

To recognize both single- and multi-user activities, they propose a novel pattern based approach.  Both activity models leverage on a discriminative knowledge pattern that describes significant differences between two classes of data. In addition, the multi-user model is able to capture user interactions. The pattern-based approach has the following advantages. First, the models for complex activities (i.e., interleaved and concurrent) can be built directly from the model for sequential activity without the training process. Second, since a discriminative pattern captures significant differences, random noise can be easily eliminated. Third, the pattern-based approach offers better scalability as compared to traditional machine learning approaches. With the above characteristics, they are moving a big step towards real-life activity recognition systems. Their comprehensive experimental results demonstrate that their approach outperforms the state-of-the-art schemes in terms of accuracy, applicability, scalability and robustness.

In addition, they also develop a hierarchical model for real-time activity recognition which first identifies simple gestures at a sensor node level, and then recognizes activities at the portable node level. There are several conference and journals publications out from these projects. The dataset and code are also available for research purpose. You may refer to this link at http://www.imada.sdu.dk/~gu for more information. 

 

Structural Health Monitoring

Part 2 – Software

 
The Illinois Structural Health Monitoring Project (ISHMP) introduced earlier is working to develop an inexpensive means for continuous and reliable SHM (Structural Health Monitoring) using dense arrays of wireless smart sensors. Researchers have designed, developed, and tested sensors to produce the high-fidelity data required for SHM that can be manufactured very cheaply. Their research has also produced an enabling customizable software framework that greatly reduces both the complexity and time for development of SHM applications for smart sensor platforms.

The ISHMP Services Toolsuite provides an open-source software library of customizable services and examples of SHM applications utilizing wireless smart sensor networks (WSSNs). The ISHMP provides users with a collection of guides to help those who are new to Crossbow’s Imote2 platform, its operating system or programming languages.

As with many wireless sensor platforms, the ISHMP Services Toolsuite on the Imote2 employs TinyOS as the operating system. TinyOS is tailored to the specific constraints of sensor network applications and occupies a small memory footprint, while efficiently supporting complex programs. TinyOS applications are implemented in NesC, a C-like programming language which supports the concurrency model of TinyOS. While TinyOS has been adopted by many sensor network applications and has quite a large user-community, it can be a daunting undertaking for engineers lacking such specific programming experience to develop code for their applications. To alleviate this problem, the ISHMP Services Toolsuite was designed assuming that the typical user will have limited background in computer science. Rather, the user is expected to be familiar with basic Windows operation and have a desire for a smart SHM system that can be installed easily and operated reliably.

 
The ISHMP Services Toolsuite employs a Service-Oriented Architecture (SOA) that lends itself to further expansion, customization, and development of WSSN applications for SHM. It provides complete applications that facilitate common tasks throughout the design, testing, deployment, and monitoring of the SHM system, while utilities offer a set of basic testing and debugging commands to be included with existing applications. The SHM Services Toolsuite includes utilities for resetting nodes remotely, listing the nodes within communication range of the local node, testing radio communication performance, and changing the radio channel and power for local and remote nodes. Specific services and tools currently available include the following:

Foundation Services: The foundation services provide commonly used wireless sensor functionalities that are required to support higher-level applications. These services include basic communication and sensing functionalities.

• Unified Sensing is a TinyOS-based sensing interface for Imote2 that supports various sensor boards including Crossbow ITS400 and Illinois SHM-A. Unified Sensing service provides precise, variable-resolution time stamping of sensor data that can be used to synchronize sensor data from different sensor nodes.
• Time Synchronization is a network-wide service for synchronizing the local clocks that sensor nodes in the network locally have.
• Reliable Communication ensures reliable data communication in a wireless sensor network. Data loss, a common problem in the wireless communication, is handled by the Reliable Communication service.
• Remote Command provides an efficient means for nodes to interact with each other. A command message is delivered to receiver nodes that process the designated tasks, and returns the list of responsive nodes and requested data to the sensor node. Remote Command allows the fault tolerant features to be easily implemented in sensor applications.
• Reliable Multi-Hop Communication provides the beta-implementation of any-to-any reliable multi-hop routing for sensor network applications. Multi-hop communication is essential for monitoring of large structures, where all sensors cannot be deployed within direct communication range of a small number of base stations.

Application Services: The application services provide the numerical algorithms necessary to implement SHM applications on the Imote2s and may also be used independently.

• Synchronized Sensing resamples the sensor data, providing synchronized data. Although the local clocks are synchronized by the Time Synchronization service, the sensing start time and the sampling rate vary from sensor to sensor with a certain amount of error. The unsynchronized sensor data is synchronized by retroactive resampling using the SyncSensing service.
• CFE estimates correlation functions between two arrays of synchronized data.
• CPSD estimates cross power spectral density between two array of synchronized data.
• RD performs the Random Decrement method to estimate unscaled impulse response functions.
• ERA estimates modal properties (i.e., natural frequency, damping factor, and mode shapes) using the Eigensystem Realization Algorithm (ERA). NExT allows the correlation functions estimated from the CFE service to be used in the ERA service.
• SSI performs the covariance-driven Stochastic Subspace Identification (SSI) for modal property estimation.
• FDD performs the Frequency Domain Decomposition (FDD) algorithm for modal property estimation.
• SDLV performs the Stochastic Damage Locating Vector (SDLV) method for damage localization.
• SDDLV performs the Stochastic Dynamic Damage Locating Vector (SDDLV) method for damage localization.

Tools and Utilities: The tools and utilities provide network testing and debugging capabilities that are necessary in any large-scale or long-term WSSN deployment. These tools facilitate evaluation of the network conditions at the structure to determine appropriate values of adjustable system parameters, and assess power consumption and longevity issues.

• LocalSensing collects sensor data from the single Imote2 connected to the PC. This tool is useful for testing new sensor board hardware as well as driver software.
• imote2comm is a terminal program that provides an interface between the PC and the Imote2 through the Crossbow IIB2400 interface board. This tool is frequently used to run the application in the ISHMP Toolsuite.
• TestServices combines numerical services, CFE, ERA, and SDLV, and performs damage detection from user-defined acceleration signals. TestServices is an application example that illustrates how the numerical services can be combined.
• TestRadio allows testing the raw bidirectional communication between a local node and remote nodes.
• RemoteSensing is an implementation of the centralized data acquisition approach that collects synchronized sensor data from a network of sensors to the base station. Time Synchronization and Synchronized Sensing services used in RemoteSensing for the synchronized sensor data.
• SensingUnit, a service component that performs synchronized or unsynchronized sensing, is particularly useful for SHM application development
• IndependentProcessingPSD is an implementation of the independent processing that decentrally estimates the power spectral densities in each node.
• DecentralizedDataAggregation is an application for the data acquisition and processing on decentralized hierarchical sensor network.

Continuous and Autonomous Monitoring Services: These services provide for continuous and autonomous WSSN operation while maintaining power efficiency.

• SnoozeAlarm provides sleep cycle functionality, which greatly reduces long-term power consumption. Sensors sleep for a period of time and then wake up for a relatively short period, during which they can interact with the network. The duty cycle is configurable by the user.
• ThresholdSentry allows a subset of the network to act as “sentry” nodes that are awakened periodically to sense data for a short period of time, determine if an interesting event is in progress, and notify the base station.
• AutoMonitor is a high-level network management application that coordinates each of its components in response to various events. It schedules sensing, data transfer, and ThresholdSentry operation according to a user-specified policy, allowing the network to operate unattended.

In addition, a library of supporting numerical functions that are common to many SHM algorithms is provided including Fast Fourier transform (FFT), singular value decomposition, eigenvalue analysis, etc. Documentation is provided for each service and test application, giving specific requirements and formats of the inputs and outputs for the service. More detailed information regarding the service-oriented architecture of the ISHMP Services Toolsuite can be found here.

To illustrate how SOA of the ISHMP Services Toolsuite lends itself to further expansion and in the development of WSSN applications for SHM, consider the figure below, which shows how the system identification method can be swapped out in an SHM application. In keeping with the SOA framework, these interchangeable services share the same input and output parameters.

 
In its current form, the ISHMP Services Toolsuite requires application programmers to simply provide the code necessary to interconnect the services and tools in a way that makes sense for their applications. In general terms, this code serves the following functions:
        • Run service X at node A
        • Send a control message to node B to run service Y
        • Send results from B to A
        • Run service Z on node A taking as inputs the outputs of services X and Y

Full-scale Integration
As described in the initial post, the software and hardware developed in this research has enabled the deployment of the smart SHM system on the Jindo Bridge in South Korea. This effort is part of a trilateral collaboration between the USA (University of Illinois at Urbana-Champaign), South Korea (Korean Advanced Institute of Science and Technology, KAIST), and Japan (University of Tokyo).

In total, 70 Imote2 sensor nodes with SHM-A sensor boards have been installed on Jindo Bridge. This SHM system constitutes the largest deployment of wireless smart sensors for civil infrastructure monitoring to date and demonstrates the suitability of the Imote2 smart sensor platform, the SHM-A sensor board, and the ISHMP software for full-scale, continuous, autonomous structural health monitoring. The figure shown below shows the measured data from WSSN on the Jindo Bridge.

 
The figure below compares computational model predictions (inset) with the experimentally determined frequencies and mode shapes obtained using the frequency domain decomposition method. As can be seen, the agreement is excellent.
 

In summary, the ISHMP Services Toolsuite provides an extensive set of middleware services, utilities, and tools that enable creation of complex, data-intensive WSSN applications for structural health monitoring. This SOA-based approach creates a framework which allows application programmers to more easily create applications for SHM systems. As a result, the framework allows more researchers, and ultimately application engineers, to design and implement successful SHM systems without the requirement of understanding how the underlying middleware and numerical services are implemented.

For information on installing the Toolsuite, visit the ISHMP site here.
For more details on the Imote2 platform, visit MEMSIC site here.

The Illinois SHM Project Services Toolsuite software is open source, but not public domain. The license, a copy of which is also included in the software distribution, allows for unlimited use and modification of the software for non-commercial purposes. Unless otherwise stated, the software and source code in the package is licensed under the OSL Software License (see LICENSE.txt), and is free for non-commercial distribution. Also included are modified files from TinyOS (BSD and Intel Open Source licenses) and LAPACK 3.0 (no formal license, free for redistribution); these licenses are indicated at the top of the respective files.

 

Structural Health Monitoring

Part 1 Hardware

Investment in civil infrastructure in the United States alone is estimated to be $20 trillion. Moreover, annual costs amount to between 8-15% of the GDP for most industrialized nations, and this investment is only likely to increase. Much attention has been focused in recent years on the declining state of the aging infrastructure in the U.S., such as the nation's bridges, highways, and buildings. The ability to continuously monitor the integrity of civil infrastructure in real time offers the opportunity to reduce maintenance and inspection costs, while providing increased safety to the public.

Bridges account for a large part of the capital investment in constructing road networks and represent a key element in terms of safety and functionality of the entire highway system. As such, informed bridge management is important to provide the public with timely transportation while maintaining a high level of safety. Manual inspection of bridges costs millions of dollars, is relatively unreliable, and can only be carried out sporadically. Recent years have seen tremendous attention directed toward development of structural health monitoring strategies, with the primary outcome sought being enhanced safety and reliability, with reduced maintenance and inspection costs. Many of the recently constructed bridges have in-depth monitoring systems; for example, the Stonecutters Bridge in Hong Kong which has over 1200 sensors. However, the enormous expense of installing traditional monitoring systems has significantly limited deployment. For example, the total system cost of the monitoring system (including installation) on the Bill Emerson Memorial Bridge in Cape Girardeau, Missouri, USA is approximately $1.3M for 86 accelerometers, which makes the average installed cost per sensor a little over $15,000. Average per node cost for other wired bridge monitoring installations are of a similar magnitude.

The Illinois Structural Health Monitoring Project (ISHMP), based at the University of Illinois at Urbana-Champaign (UIUC), is working to develop an inexpensive means for continuous and reliable SHM (Structural Health Monitoring) using dense arrays of wireless smart sensors. Researchers have designed, developed, and tested sensors to produce the high-fidelity data required for SHM that can be manufactured very cheaply. Their research has also produced a customizable software framework that simplifies the development of SHM applications for smart sensor platforms. In combination, their sensors and software create an integrated framework that can easily be utilized by most civil engineers without requiring an extensive background in computer science!

ISHMP researchers have focused on Crossbow’s Imote2, because they believe that it to be the only commercially available wireless sensor platform that can meet the high-data throughput demands of SHM applications. The powerful processor and the onboard memory of the Imote2 are the features that set it apart from other wireless sensor platforms and allow its use for the high-frequency sampling required for dynamic structural monitoring. The Imote2 has a low-power X-scale processor (PXA27x) with variable processing speed to optimize power consumption. Moreover, it has 256 KB of integrated RAM, 32 MB of external SDRAM, and 32 MB of flash memory.

The Structural Health Monitoring Accelerometer (SHM-A) sensor board was designed to meet the specific needs of SHM by Prof. Jennifer A. Rice from Texas Tech University as part of her PhD research at UIUC. The board provides flexible and accurate user-selectable sampling rates and anti-aliasing filtering capabilities to account for the local nature of structural damage where higher mode responses of the structure are often required (up to 500 Hz) in addition to low-frequency signals (DC to 20 Hz). To avoid potential signal errors, especially in the higher frequency range, minimizing sample-rate fluctuation (jitter) is critical. While simply interfacing an analog accelerometer with a high-quality ADC could address the sampling rate issues, a programmable signal conditioner was chosen because of the flexibility it provides the user in terms of anti-aliasing and signal processing. The key component of the SHM-A board is the Quickfilter QF4A512, a versatile, 4-channel ADC and programmable signal conditioner with user-selectable sampling rates and programmable digital filters. The board interfaces with the Imote2 via SPI and I2C I/O and has a 3-axis analog accelerometer for vibration measurement, one general analog input, as well as digital temperature, humidity and light sensors. One purpose of the temperature sensor is to provide onboard temperature compensation to the acceleration data. The SHM-A board provides the enabling technology that had been previously lacking to perform structural health monitoring using the Imote2.

Full-scale Implementation

After much testing and lab work, ISHMP researchers moved forward on the full-scale monitoring of the Jindo Bridge in South Korea in June. The Jindo Bridge is Korea's first cable-stayed girder bridge. A joint project between University of Illinois at Urbana-Champaign, KAIST in Korea, and the University of Tokyo in Japan, this deployment constitutes the first dense deployment of a WSN on a cable-stayed bridge. Installed with 70 nodes, the deployment is largest of its kind for civil infrastructure to date.

Prior to deploying the smart sensor network on the Jindo Bridge, the Imote2 and sensor board stacks were housed in enclosures that provided protection against environmental conditions such as rain, wind and dust. The enclosures had a gasketed lid and the ability to mount to the structure. Several nodes were designed to accommodate larger battery sources, while some were designed to utilize integrated solar panels.

Smart sensing technology for structural health monitoring is an emerging field that, by combining civil engineering knowledge with developments in sensor technology, information management, and networking technologies presents a solution that is a robust, significantly lower-cost (~$500/node), alternative to traditional structure inspection techniques. This process of implementing a damage detection and characterization strategy for engineering structures has changed the idea of SHM. The SHM process involves the observation of a system over time using periodically sampled dynamic response measurements from an array of sensors, the extraction of damage-sensitive features from these measurements, and the statistical analysis of these features to determine the current state of system health.

For more details on the Imote2 platform visit Crossbow's site here.

For more information on the SHM sensor board developed at UIUC visit the ISHMP site here and view a detailed report on the development and implementation here.

Stay tuned for Part 2 of the Jindo Bridge Deployment related to the software strategy implemented...

Testbed Testing

Wireless sensor network testbeds are critical for understanding and meeting the technical challenges of networks deployed in real world scenarios. Hardware and software testbeds have become the preferred basis for experimenting with embedded wireless sensor network applications. They provide a means for developing and evaluating sensor network technology in a controlled and instrumented environment. Experimentation with current hardware and software platforms, allows users to not only demonstrate applicability in real environments but also to validate prototypes. Compared to field deployments, the testbeds yield substantial efficiency in instrumenting potentially long-lived experiments, which is valuable in the debugging, validation, and integration phases of reliable wireless sensor networks. Universities and labs across the world have set up networks of hundreds of nodes using a Mote platform from Crossbow's suite of wireless sensor network devices choosing from simple platforms such as the TelosB to advanced devices like the Imote2.

Researchers at the University of Buffalo, SUNY and Georgia Institute of Technology have written an article focused on the the idea of taking wireless testbeds to the next level by incorporating multimedia and characterizing the challenges of wireless multimedia sensor networks (WMSNs). The availability of low-cost hardware has enabled the development of WMSNs, i.e., networks of resource-constrained wireless devices that can retrieve multimedia content such as video and audio streams, still images, and scalar sensor data from the environment, along with standard wireless sensor networks. Research is being conducted on prototypes of multimedia sensors and their integration into testbeds for experimental evaluation of algorithms and protocols for WMSNs. Open research issues and future research directions, both at the device level and at the testbed level, are discussed and tested constantly.  Network testbeds allow the effectiveness of algorithms and protocols to be evaluated by providing a controlled environment for measuring network performance.

Every testbed utilizes a specific Mote platform that is optimized for that particular testbed's focus. The wireless sensor platform is chosen based on its available capabilities to allow for multimedia integration. A WMSN is a distributed wireless system that interacts with the physical environment by observing it through multiple media. Furthermore, it can perform online processing of the retrieved information and react to it by combining technologies from diverse disciplines such as wireless communications and networking, signal processing, computer vision, control, and robotics. Applications in the real world that would benefit from WMSNs were categorized into the categories of surveillance, traffic monitoring and enforcement, personal and health care, gaming, and several elements of environmental and industrial monitoring by researchers at SUNY, Buffalo and Georgia Tech. Testbeds allow the observation of the performance of the WMSN in a controlled environment. Hence, the effect of different types of inputs, physical operating conditions, and subjects for sensing can be studied, and the functioning of the devices in the testbed may be changed appropriately for accurate measurement. A few WMSNs developed are outlined below:

A visual sensor testbed was developed as part of the Meerkats project to measure the tradeoff between power efficiency and performance. Results revealed that there was significant energy consumption in keeping the camera active, and writing the image to a Flash memory followed by switching the camera off conserved energy. There was also a finite instantaneous increase in the energy consumption due to state transients. Researchers were also able to determine that the processing-intensive benchmark resulted in the highest current requirement, and transmission was shown to be only about 5% more energy-consuming than reception.

Expandable, vision-, and sensor-equipped wireless robots with MICA sensor motes for networking were designed in the Explorebots testbed architecture. The target localization experiments on the testbed, composed of these mobile robots, used on board multimedia sensors such as custom-designed velocity and distance sensors, motor movement control, an in-built magnetic two-axis compass, and sonic sensors. By processing the sound and light sensors outputs, the robots were guided towards the target source.

The Mobile Emulab network testbed provided a remotely accessible mobile wireless and sensor testbed. Robots carried motes and single-board computers through an indoor field of sensor-equipped motes. A remote user could position the robots, control all the computers and network interfaces, run arbitrary programs, and log data in a database. The path of robots, which was also equipped with Webcams, could be planned, and a vision-based system provided positioning information with an accuracy within 1 cm.

IrisNet (Internet-scale resource-intensive sensor network services) was an example software platform for a heterogeneous WMSN testbed. Video sensors and scalar sensors were spread throughout the environment and collected potentially useful data. IrisNet allowed users to perform Internet-like queries to video and scalar sensors that spread throughout the environment. The architecture of IrisNet was two-tiered: heterogeneous sensors implemented a common shared interface and were called sensing agents (SAs), while the data produced by sensors was stored in a distributed database that was implemented on organizing agents (OAs). Different sensing services were run simultaneously on the architecture. For example, a set of video sensors could provide a parking-space finder service, as well as a surveillance service.

In the design and implementation of SensEye, a multiple-tier network of heterogeneous wireless nodes and cameras, was created to test surveillance applications. Each tier comprised nodes equipped with similar cameras and processing ability, with increasing resolution and performance at each stage. The lowest tier consisted of low end devices, i.e., MICA2 Motes equipped with 900 MHz radios interfaced with scalar sensors, e.g., vibration sensors. The second tier was made up of motes equipped with low-fidelity Cyclops or CMUcam camera sensors. The third tier consisted of Stargate nodes equipped with Webcams that could capture higher fidelity images than tier 2 cameras. Tier 3 nodes also performed gateway functions, as they were endowed with a low-data-rate radio to communicate with motes in tiers 1–2.The aim was to efficiently undertake object detection, recognition and tracking by triggering a higher tie into the active state based on a need basis.

The WMSN-testbed at the Broadband Wireless Networking (BWN) Laboratory at Georgia Tech was based on commercial off-the-shelf advanced devices and had been built to demonstrate the efficiency of algorithms and protocols for multimedia communications through wireless sensor networks. The testbed was integrated with the scalar sensor network testbed, which was composed of a heterogeneous collection of Imote and MICAz motes from Crossbow. The testbed allowed the integration of heterogeneous devices in experimental testbeds and some succesful examples in developing APIs and system software for WMSNs.

These are just a few of the various wireless sensor network testbeds found worldwide. Crossbow's vast portfolio of wireless sensor platforms provides researchers and government/commercial users with the equipment they need to set up a lab for their wireless sensor course or to verify their specifications prior to real world deployment of their wireless sensor network. For information on how to set up your own WSN testbed or for details on Crossbow's wireless sensor network platforms, contact sales@xbow.com.

GOOOOOOAAAALLLLL!!!!

Soccer (or as it is more widely known - football) has often been considered the most popular sport in the world. Wherever you go, there are kids on the street or in a park kicking around a ball, jerseys from the most popular football clubs can be seen in every major city, and the truest World Cup truly does include most of the nations on this earth. It seems fitting that Motes would enter the arena of sports through soccer. The idea behind Motes and wireless sensor networks is to create a device that will become ubiquitous throughout this planet enabling us to live in a smarter more connected world. Soccer has connected people from all over, at all ages with all levels of skill for generations, and now Motes have made their way into this arena with a little bit of fun!

Embedded systems are increasingly becoming connected through wireless networking.  These devices now form the basis of many of today’s consumer products including cell phones and video game controllers.  In the CSE466 “Software for Embedded Systems” class in the Department of Computer Science & Engineering at the University of Washington, students used the design of a multi-player video game as motivation for the  principal concepts in wireless embedded systems.  Each student in the class designed an accelerometer-based game controller and then, the class as a whole, developed a multi-player video game that allowed 28 players (the number of students in the course) to play simultaneously. How did they do it? With Motes!

Computer Engineering curricula have traditionally included the interfacing of sensing and actuation devices to microcontrollers but have not emphasized wireless communication. This time students at University of Washington undertook the task of updating the platform used in this course to Crossbow's Imote2 which runs an embedded Linux operating system and developing a multi-player video game using controllers modeled on the, then just introduced, Nintendo Wii video game controller. Previous courses had used the MICA Mote platform, but the Imote2 offered a device with greater capabilities.

Each student designed their own Wii-like accelerometer-based 2-D game controller.  An LCD screen was added so that game state could be displayed on the user’s controller during the game. The Imote2 platform offers users an expansion connector for attaching sensing and actuation hardware.  On one side, students attached the basic ITS400 Imote2 sensor board which includes a 3-axis accelerometer, temperature sensors, a humidity  sensor, a light sensor, and a 4-channel A/D for further additions.  On the other side, students designed their own board to provide some actuation known as the "SuperBird".  A cell phone-size color LCD screen as well as sound generation capabilities, a speaker, microphone, and audio jacks.  This board also included a cell phone camera, jog dial, USB host port, barometer, and a heart rate sensor and is the form factor of a large cell phone. While not all of the board’s capabilities were used in this particular application, the board was designed to be flexible enough to accommodate a variety of projects (such as a video phone, music player, etc.) in the future. The Imote2 platform is based on Linux and students were provided with the device drivers for their project implementation.

 

Students decided on utilizing controllers that could move a player in two dimensions.  Soccer quickly emerged as a game that could be varied to accommodate the project requirements, namely, that it should involve every student simultaneously and require only two-dimensional control of each player.  In addition, students wanted to have some collaborative elements between players that would spur real team play - soccer being a true team sport! The game was developed in steps.  Development began with the basic player controller, that is, the mapping of values from the accelerometer to an X-Y velocity vector.  A communication protocol needed to be devised on top of the basic MAC of the 802.15.4 radio to handle the communication between players and  the game coordinator.  To ensure that there was some inter-player communication, the “captain” of the merged players (the player with
the highest number) collected all the moves of the constituent players and reported that result to the game coordinator.  The scheme was basically round-robin.  The game coordinator polled the first player for its move and waited for a response before proceeding to the next player.  If a player was too slow in responding, its movement was ignored for that round.  This provided some timing constraints on the implementations to quickly handle packets from the game coordinator.  The boards were programmed to allow students to control players in the virtual soccer-like game. Tilting the board would cause the player (represented by a dot on the field) to go in the corresponding direction.

On the final day of the course a 30-minute soccer match was held between the two sections of the class.  A video of the match and an explanation of the project can be seen here:

GOOOOOAAAAAAAALLLLLL for CSE466 at University of Washington!!

Robocopters and Motes

Crossbow's Motes were featured last month in an article on IEEE Spectrum. The application featured showcases how the Motes can be used when deploying mobile distributed communications networks.

The meter-long helicopters lined up under the fluorescent lab lights at the Berlin Technical University might look like overgrown toys, but they’ve got a little more under the hood. These are flying robots. They take off, land, and scout terrain autonomously and are being wired to deploy ad hoc communications sensor networks. And unlike any other robocopters, they can work together.

Researchers expect they’ll be used to distribute sensors that would help coordinate firefighting efforts or search flood zones, to track or find people and vehicles, or to shoot movies and cover sports events. Hoisting communications gear, they could one day help channel radio, Wi-Fi, or mobile phone traffic where infrastructure has been wiped out by an earthquake or other natural disaster.

Several groups around the world are working on miniaturized robot helicopters with advanced intelligence, notably in California, South Carolina, the Netherlands, South Korea, and the UK. But the Berlin team believes it is the first to write software and build systems that get multiple robot copters to collaborate. The project brings together a half-dozen institutions and companies from across Europe.

The copters’ control systems allow small craft to work together in lifting loads and scouting environments. Coordinated, the copters can lift weights that would normally require larger, exponentially more expensive machines. Estonian robotics engineer Konstantin Kondak, a professor at TU Berlin and one of the project’s leaders, says that having three or four copters in the air, each sharing the load while tethered by a rope to a single object, creates too many contrary forces for a set of human pilots to handle: “If you try to do this flying manually, it is not a stable system. You have to correct at all times; it’s too much.” But autonomous robots, making instant and coordinated adjustments, can do the job, he says.

Each robot must account for the location of the other helicopters, the forces coming from them, and the load on the rope, to jointly lift something. The helicopters’ coordination comes from a system that integrates four software modules for stabilizing the copter: one for navigation, one for exploration, one for obstacle avoidance, and one for processing orientation, horizon, and position.

The robocopters are good for much more than just lugging things around. Project manager Aníbal Ollero, a professor of engineering and automated systems at the University of Seville, in Spain, says that a flexible, easy-to-transport team of choppers makes for more efficient scouting because they automatically divide an area among themselves.

They’re also faster at another important task: deploying distributed communications networks by dropping off sets of tiny sensor nodes. These nodes combine a data processor, a radio, a battery, and—depending on what needs measuring—temperature, light, radiation, location, or chemical sensors. For the autocopter project, off-the-shelf nodes from wireless-sensor firms Crossbow Technology and Ambient Systems were optimized and linked by data-routing experts at universities in Stuttgart, Germany, and Twente, Netherlands. Just a few centimeters across and 7 millimeters thick, the individual nodes transmit over a range of just 25 meters, but as a network they pass radio messages to one another to get to a central unit (or hovering robocopter). Hundreds of thousands of these nodes could be distributed by robot to survey a forest fire or flood zone for rescue efforts, according to the researchers.

Final trials for the project get under way, far from the Berlin winter—this spring in southern Spain, Ollero says. If all goes well, helicopters will deploy a sensor network, track mobile objects and people, follow movement inside and outside of buildings, and capture it all with high-end airborne cameras.

For more information on Crossbow's Mote platforms, visit www.xbow.com.

Sensors Help Keep the Elderly Safe, and at Home

The New York Times published a front page story regarding how sensor technology is changing the face of home health care. The article discussed how increasingly, many older people who live alone are not truly alone. They are being watched by a flurry of new technologies designed to enable them to live independently and avoid expensive trips to the emergency room or nursing homes.

Crossbow's wireless sensor network technology has been used in a plethora of various research and project deployments on the capabilities of integrating wireless sensor network technology for elder care. Recently, researchers at Kalasalingam University in Nadu, India published their project details in the Jan/Feb Issue of Telemedicine and eHealth on using the IRIS and Imote2 platforms for home health monitoring. The issue they address is real as the increasing demand on public health care due to the aging population has become a major problem in developed countries. With the increasing number of elders relying on home care, better monitoring and analysis systems are crucial for maintaining and improving the quality of life for the elder patients. The concept of health monitoring is advanced by which health parameters are automatically monitored at home without disturbing daily activities. The proposed system is a network that supports various wearable sensors and contains on-board general computing capabilities for individual event detection, alerts, and communications with various medical informatics services. The purpose of their system is to provide extended monitoring for elderly patients under drug therapy after infarction, data collection in some particular cases, and remote consultation for elderly people.

With the advancement in ubiquitous computing techniques, researchers at Kalasalingam University proposed this new system for health monitoring at home to assess the elderly peoples’ health status. The integration of sensing, information, and communication technologies allowed elderly people to be constantly monitored. Moreover, constant monitoring would increase early detection of adverse conditions and diseases in elderly patients, potentially saving more lives. The proposed healthcare system was built upon a global medical information system made up of three main components: (1) Various sensors of different types to monitor physiological signals of elderly patients, their environments, and their activities with sensor nodes kept in the patients’ home; (2) Sensor nodes to transmit the signal from sensors to central servers through a base station; and (3) Central server for management of a knowledge database related to patients and responsible for broadcasting messages and alarms to healthcare professionals and caretakers. The system proposed a technique for monitoring elderly patient status at home and detecting critical situations for the purpose of alerting the doctors as well as the caretakers. It consists of a front end, the patient station, and the server.

The Patient Station, consists of an IRIS (XM2110CA) sensor node, Crossbow's newest 2.4GHz Mote platform for ultra low-power, long-range wireless sensor networking that receives information from the physiological sensors. The 3D camera kept at home was connected with the Imote2 platform (IPR2400CA) to visually monitor the elderly patient in the home environment in addition to the physiological signals. The Imote2 sensor node is aimed at applications involving data-rich computations, where there is a need for both high performance and high bandwidth, which require greater processing capability and low-power operation with a low duty cycle to achieve longer battery life. Sensor nodes are responsible for sensing as well as the first stage of sensory data processing in the data communication. The received signal from the patient station was transmitted to the central server for analysis. Researchers found that the benefits of the IRIS mote included excellent radio frequency range, substantially lower sleep current, and double the program memory of other Mote offerings. By Integrating a high-performance, low-power PXA271 Intel processor and an 802.15.4 radio with a built-in 2.4-GHz antenna, the Imote2 provided a platform for digital imaging applications.

Researchers placed sensor nodes equipped with flex sensors on the chest for ECG, and on the right hand for heart rate and acceleration measurement. The IRIS sensor node placed on the body was responsible for collecting physiological data and movement data from the sensors and transmitting it to the central servers through the aggregation node. The aggregation sensor nodes do not act merely as data collecting and forwarding points. Instead, they make decisions whether the information should be stored for future use, relayed as they are, or modified by applying computation and aggregation with other data.

The server, which is the core processing element, received the patient’s data from the patient station through the base station and stored the data into the patient database, while performing long-term trend analyses and prediction. It dispatches the critical events detected by sensor nodes to healthcare professionals. Moreover, it coordinates and controls the overall functionality of the system. This system can also give the alert to the doctors via a personal digital assistant (PDA) or cell phone. The sensor nodes could be programmed to awaken the node whenever an abnormal signal is detected and transmit the data to the server and then return to sleep mode. When an abnormality was detected in the sensor node, it would automatically turn on and transmit the signal to the central server and simultaneously alert the caretaker near the patient. The server, which is the core processing element, received the data regularly from the sensor nodes for analysis. The analysis was done using Crossbow's MoteView, and operates under TinyOS, which offers a built-in library for data acquisition, processing, analysis, and display.

The central servers were programmed with two algorithms. First was the threshold based algorithm, which attempted to identify the physiological parameter values that were potentially harmful or indicative of immediate danger to the patients. The algorithm detected the upper and lower threshold values from sensors output and alerted medical personnel and caretakers when the patient was not physically capable of requesting help. Second was the inactivity detection algorithm for detecting rapid movement or lack of movement of elder patients from accelerator output. The sensor node was programmed through the central server in such a way that if lack of movement/ action is detected, it awakened the Imote2 device connected to a 3-D camera fixed in that specific room to send the status of the elder patient to a central server. Upon detecting an anomalous event, both algorithms would sound an auditory alarm from the central server to emergency departments and alert the caretakers near the patients.

In this implementation, researchers at Kalasalingam University were able to connect two different systems, that is, the patient record database and the Web portal, through the use of well-defined Web services. Patient information was transmitted over SOAP, a secure and encrypted form of XML. The WSDL (Web Service Definition Language) for these Web services is published to a community of authorized users. This Web-service based approach for inter-system communication gives the system the flexibility to operate with third-party software in the future.The proposed system used a wireless sensor network technology within the residence in order to increase functionality, security, and quality of life.

As the Times put it, 'The future of these technologies, and the terabytes they gather, can involve unprecedented information about the whereabouts and well-being of older people.'

IMB400 Imote2 MultiMedia Board Wins EG3 Editor's Tech Choice Award

Crossbow's IMB400 Imote2 MultiMedia board has been chosen as the recipient of the EG3 Editor's Tech Choice Award for Winter 2009! The platform is now in the running for the Readers Choice Award! The eg3 Tech Choice awards are the newest and one of the largest "reader's choice" awards in the embedded and electronic design industries. The awards commend new products that take promising technologies and provide practical ways that design engineers and programmers can turn these technologies into real designs, today.

To participate and vote for the IMB400 in the Reader's Choice Awards, please visit the eg3 site here and take their survey to vote for your choice in the different product categories. Crossbow's Imote2 Multimedia Board (IMB400), an integrated camera sensor board, is up for the 'Best Wireless Product' Reader's Choice Award. The IMB400 is innovative in the way it simplifies the capture of rich media content for wireless sensor network applications. Building on the popular Imote2 advanced wireless sensor platform, the IMB400 offers a compact, power efficient solution due to its integration of camera, audio and motion detection functionality into one platform. The built-in camera can handle high-quality images with resolutions up to 640x480 pixels and 30 fps, along with audio at sampling rates of up to 48kHz.. For more information on the IMB400 Imote2 Multimedia board and to purchase this platform, visit Crossbow's site here.

eg3.com is the oldest and largest web resource devoted to electronic design, with a focus on embedded systems, real-time computing (RTC), and digital signal processing (DSP). Founded in late 1994, eg3.com indexes free and non-commercial design information on the net, summarizes the over 800 vendor companies that make up this marketplace, and organizes the content of major chip companies and third party vendors into a searchable "design tool" for embedded engineers and designers.

Imote2.Net edition featured on EngineeringTV

The .NET Micro Framework is a platform that enables developers to more quickly develop embedded systems that are smart, securely connected and easier to manage. Jonathan Kagle, Microsoft's .NET Micro Framework Senior Program Manager, introduces some of the hardware developers are using in Micro Framework devices, including Crossbow's Imote2.Net Edition wireless sensor node platform on Engineering TV.

The Imote2.NET Edition (IPR2410) is built around the low-power PXA271 XScale processor and integrates an 802.15.4 radio (CC2420) with a built-in 2.4GHz antenna. The Imote2 IPR2410 is factory configured to run .NET Micro Framework. It is also sold as part of the Imote2.Builder kit.The Imote2.NET Edition is a modular stackable platform and can be expanded with extension boards to customize the system to a specific application. Through the extension board connectors sensor boards can provide specific analog or digital interfaces. A battery board is provided to supply system power, or it can be powered via the integrated USB interface.

Open SESAME!

These magical words won't take you to a cave of treasures, but they do lead us to some exciting applications for wireless sensor networks that open up a treasure trove of information previously inaccessible. SESAME: SEnsing for Sport and Managed Exercise has a vision in which athletes and coaches are continuously provided with precise and relevant information about their performance, their body state and posture, presented in a form determined by sport-specific training requirements based on a careful analysis of coaching methods and coaches’ informational needs. To realize this, athletes wear an easily-extensible range of different sensors that capture accurate information about their position, skeletal posture, muscular response, and physiology in a way that is non-intrusive and capable of working in the context in which the athlete normally performs. This setup is engineered so as not to cause injury, discomfort or performance degradation and it must not interfere with aerodynamics. Wearable sensors are complemented by track-side monitors and video capture equipment and by an integrated hardware/software/network platform designed to enable substantial volumes of data to be gathered, recorded, analyzed and presented to athletes and coaches in the most accessible and useful form. The main objective of the SESAME project is to conduct high-quality scientific research to produce deployable systems that have a positive and measurable impact on the training of elite athletes.

Researchers at the University of Cambridge have just released an open source low power 802.11 sensor board tailored for Crossbow's Imote2 platform. The SESAME consortium is a multidisciplinary group consisting of 6 partners with University of Cambridge Engineering Department as a member. The sensor board developed is designed to support the high data rate requirements of the sports sensing application. The UCAM-WSB100 is a low power wireless sensor daughter board for the Imote2 platform designed to facilitate high data-rate wireless sensor applications. Developed to support the SESAME project, the board will be used to collect real-time and off-line processing and feedback in enhancing the performance of elite athletes. 

Overview of the UCAM-WSB100 sensor board:
* Compatible with the Imote2 processor board
* Low Power 802.11 b/g based on Marvell 88W8686 chipset
* Supports ad-hoc and infrastructure modes
* Access to power control of the radio system
* 12 analog channels (12-bits resolution)
* Physical dimensions: 48mm x 36mm
* Linux driver support via libertas drivers in mainline kernel

 

The information obtained from the sensors will be pre-processed on the athlete to take account of the measurements required and the prevailing network conditions. The data is then transmitted wirelessly to a base-station and application platform. Here, the data will be further fused and processed in a way that is informed by an understanding of biomechanical models of athletes; an understanding of the consequences of sensor placement error and physical properties of the method of attachment; and the coaching objective for which the data are being captured. Within SESAME, the primary experimental focus will be on sprinting, for which precise technique is hugely important and mechanical constraints on performance are well understood. Consequently, the derived data will include the position, velocity, acceleration and orientation of the athlete, their stride length and rate, body posture and instantaneous pressure in the shoes, as well as physiological data such as heart rate and blood oxygen level. The data will then be output in three different ways: they will be sent for long term storage and offline analysis; they will be presented visually to the coaches in a way that is meaningful to them; and they will be returned directly to the athlete in real time as biofeedback.

The UCAM-WSB100 is created as an open source platform. As part of the effort to develop open source hardware for use in conjunction with Crossbow's Imote2, the details of the hardware reference design (schematics, board layout diagrams) are available to the wireless sensor network community. Users will be able to easily tailor the system to their particular application. Researchers at the University of Cambridge are seeking volunteers to help improve the hardware reference design and add support for this board in other operating systems such as TinyOS and .Net. If interested in participating in this development, click here and for more information on this project, click here.