INTRODUCTION

The development of sensor network technology is of great national importance

by virtue of critical applications directly related to the Australian environment. Sensor
networks provide the ability to gather accurate and reliable information, enabling early
warnings and rapid coordinated responses to potential threats. This encompasses the
ability to enhance national security from hostile threats as well as the ability to save lives
through environmental monitoring of natural disasters. Environmental sustainability can
also be improved through sensor network monitoring, by protecting valuable resources
from overuse or damage, as well as being able to collect valuable information previous
considered too difficult and too costly.

Applications of distributed sensor networks in habitat monitoring serve to provide

An Environmentally Sustainable Australia , by offering the potential for cost effective
solutions for monitoring environmental conditions. The feedback obtained from sensor
networks, provide valuable scientific information enabling a greater understanding of the
environmental impact of development, climate change and water usage for domestic use
as well as farming. In addition, the data collected by the appropriate sensors can be
applied directly for irrigation control, thus supporting Australian industries while
ensuring that steps are taken to implement sustainable practices. Further information may
be extracted pertaining to soil conditions, assisting in agriculture management for the
prevention of soil degradation. Such is the potential sensor development that the
monitoring and understanding of delicate eco-systems can be a viable proposition where
it is now often prohibitive. The potential cost effectiveness of intelligent sensors may
allow for large scale monitoring of emissions and the subsequent modification of
industrial processes. . One of the key research themes of the network is based on the
principle of smart information use. Efficient data management is a fundamental problem
faced in sensor network development in terms of the volume of data (particularly in
imaging), but also in terms of the value of the associated information. The outcomes of
this work will not only advance sensor network development but also related
applications.

1
ABOUT SENSOR NETWORKS:

Recent emergence of very large number of diverse sensors and sensor networks
has the potential to impact on the quality of all areas of life. Scientific challenges in
realizing this potential is significant because of the multidisciplinary nature and
complexities involved. This research network builds on the best scientific talent
available in the interdisciplinary areas (biology, mathematics, statistics, computing,
electrical engineering and mechanical engineering) with the best overseas scientific teams
to solve the underlying scientific problems to enable the Australian industry to exploit
and apply this technology in areas of defence, health care and environment.

There are various focus of the sensor networks

1. Micro and Nano Sensors

2. Distributed Sensor Networks
3. Surveillance and Monitoring
4. Sensor Fusion and Tracking
5. Scheduling and Optimisation
6. Machine Intelligence

A unique feature of sensor networks is the cooperative effort of sensor nodes fitted
with onboard processors. Monitoring of activities overseas, bush fires, health, homes,
streets, airports, ports and the internet will be revolutionized using multiple interlinked
cooperative sensors enabling coordinated and timely response systems. Researchers of
this initiative are recipients of ARC special research centers, international grants,
federation fellowships and other major international initiatives focusing on aspects of the
central theme of this Network. This initiative facilitates to bring these isolated efforts to
fulfill the vision of the network. The aim of the sensor networks is the new initiative will
address the constrained multi-sensor problems to develop intelligent, economically viable
solutions of value to defence, homeland security, health sciences and environment.

2
 By drawing inspiration from biological multi-sensor systems and practical
insights from their artificial counterparts, this initiative will address the fundamental
problems relevant to the central theme of the network with an interdisciplinary approach.
By drawing inspiration from biological multi-sensor systems and practical insights from
their artificial counterparts, this initiative addresses the fundamental problems relevant to
the central theme of the network to develop intelligent, economically viable solutions of
value to defence, home land security, health sciences and environment. Specifically,

1. Australia to be the leading country in emerging field of intelligent sensor networks.

This has a huge potential for spin-off serving both local and global markets in the
business of defence, homeland security, bio-medical and environmental protection.
2. Development of Cost-effective, scalable surveillance solutions using unmanned
aerial vehicles with ad hoc sensor networks for defence and home land security
enhancing the security capabilities of Australia.
3. Development of Revolutionary networked bio/nano sensors integrated with living
organisms for breakthrough applications in medicine further advancing the
Australia's leading position in health care industry.
4. Early disaster monitoring e.g., bush fires and other sensitive environments of
immense value to Australia, using large number of cheap, geographically
distributed sensors with limited communication capabilities saving billions of
dollars in damages.
5. A unique, leading research network with distributed resources leveraging
off high priority research initiatives in the field of intelligent sensors, networks and
information processing will be established.
6. By enabling a closer interaction between graduate students, post doctoral
fellows, leading researchers, and industry representatives, it produces a new
generation of entrepreneurs creating wealth for the nation.

3
 RESEARCH PROGRAMS

The following research themes emerged after 3 successful meetings with experts in
different field over a period 2 months. The list covers most of the major areas in the field
but is incomplete. A couple of theme areas would be added in the due course.

• Intelligent Sensors
o Nanotechnology-Enabled Sensors
• Sensor Networks
o Sensor Scheduling and optimisation
o Data Fusion & Tracking
• Information Processing
o Surveillance and Monitoring
o Machine Learning
o

4
INTELLIGENT SENSORS:

To support the requirements of distributed sensor networks, sensors must possess

greater functionality than simply gathering data and blindly transmitting the data to a
centralized sensor node. Intelligent sensors are an extension of traditional sensors to those
with advanced learning and adaptation capabilities. The system must also be re-
configurable and perform the necessary data interpretation, fusion of data from multiple
sensors and the validation of local and remotely collected data. Intelligent sensors
therefore contain embedded processing functionality that provides the computational
resources to perform complex sensing and actuating tasks along with high level
applications. The functions of an intelligent sensor system can be described in terms of
compensation, information processing, communications and integration. The combination
of these respective elements allow for the development of intelligent sensors that can
operate in a multi-modal fashion as well conducting active autonomous sensing.

Compensation is the ability of the system to detect and respond to changes in the
network environment through self-diagnostic routines, self-calibration and adaptation. An
intelligent sensor must be able to evaluate the validity of collected data, compare it with
that obtained by other sensors and confirm the accuracy of any following data variation.
This process essentially encompasses the sensor configuration stage.

Information processing encompasses the data related processing that aims to

enhance and interpret the collected data and maximize the efficiency of the system,
through signal conditioning, data reduction, event detection and decision making. This
may involve a collection of filtering and other data manipulation techniques together with
advanced learning techniques for feature extraction and classification in order to provide
the most relevant data in an efficient representation to the communications interface.

Communications component of intelligent sensor systems incorporates the

standardized network protocol which serves to links the distributed sensors in a coherent
manner, enabling efficient communications and fault tolerance.

5
 Traditional task specific sensor systems often contain a number of limitations in
terms of complexity, cost and flexibility. Intelligent sensors aim to overcome these
limitations through the utilization of standardized transducer interfaces and
communications protocols, resulting in autonomous, distributed, re-configurable sensors.

Integration in intelligent sensors involves the coupling of sensing and

computation at the chip level. This can be implemented using micro electro-mechanical
systems (MEMS), nano-technology and bio-technology. A hierarchical structure can be
used to describe the functionality of the system, where the lower layer performs the
signal processing functions, the middle layer performs the information processing and the
upper layer performs the knowledge processing and communications.

Validation of sensors is required to avoid the potential disastrous effects of the

propagation of erroneous data. This is different problem than overcoming individual
sensor failure. A control system operating decisions made on faulty data can lead to
unpredictable behaviour or even complete system failure. The impact of such errors may
be reduced through the use of a dense sensor network. The incorporation of data
validation into intelligent sensors increases the overall reliability of the system. So an
effective means for performing this function is required.

Data fusion techniques are required in order combine information from multiple
sensors and sensor types and to ensure that only the most relevant information is
transmitted between sensors. Consequently, the load on network bandwidth is kept at an
acceptable level. The area of sensor fusion can be approached from a variety of
perspectives. Biological science has been used to consider how sensor fusion is
accomplished, while cognitive science has explored why sensor fusion is an integral part
of perception.

There are various benefits for intelligent sensors, Intelligent sensors operating in a
task specific manner with effective data collection techniques enable the development
and application of more flexible sensor networks that efficiently utilize and coordinate
the limited resources of each individual sensor.

6
 By focusing resources according to the state of the surrounding environment and
on the immediate task, more efficient operation of the sensor and is ensured.

Accuracy: An intelligent sensor will incorporate features that enable it to

compensate for systematic errors, system drift and random errors produced due to system
parameters or the characteristics of the sensor.

Reliability: The incorporation of data and sensor validation techniques to detect

corrupted data, self-testing of network path connections and sensor operation, as well as
calibration of sensor drift, provides yet another level of system reliability in addition to
techniques already applied in the network design.

Adaptability: The processing parameters of an intelligent sensor system should

be determined automatically and adopted by a higher level in the system architecture.
This enables the optimization of the measuring and processing operations, as well as
enabling the sensor to adequately respond to changing environmental conditions.

The following are some existing research applications:

Automatic Assembly: Engineering components and assemblies could be made

more like organisms in there ability to self-assemble, thus having a significant impact
upon production speed, capacity and complexity. Self-repairing abilities are an obvious
side-effect of such abilities.

Microendoscopy: The ability to navigate micro or nano-structures through the

human body has the potential to make a significant impact on modern medicine.
Understanding the mobility and sensory systems of parasites, worms and insects, may
provide the necessary design information to realize this objective.

Intelligent suspension for automotive applications: A basic tuned suspension

unit requires a spring and damper unit. Tuning the suspension to a particular frequency is
generally overlooked. Vertebrate muscle maybe mimicked to construct a high
displacement and high damping spring system.

7
 Gels: The principle of a fluid enclosed in a membrane being made to do useful
work can be seen in our own muscular system, plants and in the skin of worms.
Contracting the muscles in the body wall and increasing its internal pressure the worm is
able to change shape. Controlling the swelling and contracting of a polymer gel
appropriately encased, enables a system to work as an artificial muscle.

Smart fabrics: Analyzing the insulation layers of animals and other natural
responses to temperature fluctuations may contribute to the development responsive
clothing, with properties based on the state of activity of the wearer. This would reduce
the number layers required by the wearer while remaining suitable for a variety of

8
weather conditions.

Nano technology- enable sensors:

9
 Current nanotechnology permits the operation on the scale of atoms and
molecules. This promises to have a dramatic impact on sensor design and capabilities.
Nanotechnology has become a key technology in sensor development. Sensors can now
exploit novel properties of materials at the nano-scale. Chemical and biological materials
operate at the nano-scale, hence nanotechnology is well suited to design of chemical and
biological sensors.

Initial research in nanotechnology involved ministration of the macro techniques.

The small size of these sensors leads to reduced weight, low power requirements and
greater sensitivity. In recent years there has been a large increase in the amount of
funding put into nanotechnology. Over the last 5 years there has been a 5 fold increase.
Industries befitting from Nanotechnology include transportation, communications,
building, medicine, safety and security.

The new possibilities now available are endless. The ability for atomic
bricklaying lets the designer be precise when designing new sensors. This ability will
reduce the amount of defects in new devises. At the atomic level the materials have new
properties which can be exploited like surface and quantum effects. Nanotubes have been
shown to have a number of uses in sensor technologies. They are extremely narrow
hollow cylinders made of carbon atoms. The orientation of the carbon atoms can affect
the conducting and semi-conducting properties. These can be used to integrate electrical
circuits for the design of sensors. These nanotubes can be grown on existing structures.

The existing IC technologies can be used to integrate these nanosensors into integrated
electronic circuits. The sensor chips can be used as building blocks to build new more
complex sensors.

Nanotechnology has a vast number of applications. IBM is working on data storage

proximal probes. These can make and read nanometer-scale indentations in polymers.
The current densities are around 1x1012 bits per square inch which is greater then the
current magnetic based recording devices.

10
 Advances in nano-manufacturing have been from the top-down approach and
bottom-up approach. Conventional microelectronics (lithography, etching, and
deposition) has approached the nanometer scale. The current line widths in chips are near
the 100 nm. Manufacturing from the bottom-up is also possible using individual atoms
and molecules to build useful structures. IBM has managed to write IBM using individual
atoms. Designers can also combine micro and nanotechnologies to develop new sensor
systems.

Using computers for the design of new nanotechnology is important. Understanding the
interactions of atoms and molecules is required when simulating using powerful
computers and algorithms.

Working at such a small level also has it's own problems. The new sensors are very
sensitive. These sensors are prone to degradation from the effects of foreign substances,
heat, and cold. At such a small scale the micro effects become more significant. This
problem can be partially overcome by installing hundreds of sensors in a small space.
This allows malfunctioning devices to be ignored in favor of good ones.

The amount of applications of nano-sensors can be applied to is only limited

by the imagination. Physical sensors, electro-sensors, chemical sensors and bio-sensors
can all benefit from nano-technology.

Walter de Heer has devised a Balance . This balance is the smallest in the world. A
particle to be weighed is placed on a nanotube. The mass of the particle was calculated
from changes in the vibrational resonance frequency with and without the particle. The
balance can be used to weigh signal molecules.

Measurement of electricity is important and the bases for a large number of sensors. A
submicron mechanical electrometer demonstrated charge sensitivity below a single
electron charge.

11
SENSOR SCHEDULING AND OPTIMISATION

A sensor network is an array of sensors of diverse type interconnected by a

communications network. Sensor data is shared between the sensors and used as input to
a distributed estimation system which aims to extract useful information from the
available sensor data. When all the data from all the sensors in the network is available at
one place, conventional centralized estimation techniques can be employed. Centralized
multi-sensor estimation theory critically assumes that the choice of which data to send
from any particular sensor to the centralized node is fixed. If the network bandwidth
changes during operation then such a system design can at best respond in a pre-planned
manner. The centralized approach is typically only feasible for certain non-critical static
sensor situations, as it fails to adequately address the issues of scalability or of
survivability under real world degradation. Destruction of the central node (or an
associated critical communications link) results in total network failure. A scalable
system is required, in which new sensors or network links can be easily added, that
continues to provide useful information even when some sensors and parts of the network
fail or are destroyed, and can continue to operate within performance bounds while the
communications bandwidth is varying (perhaps due to environmental conditions or
electronic jamming).

In order to realise these practical requirements, a distributed estimation

architecture as shown in Figure 1 is needed. Distributed data fusion overcomes many of
the limitations of centralized fusion but also introduces new problems. To date distributed
data fusion systems have been heavily reliant upon having very high communication
bandwidths, due to the large amount of sensor data that must be transferred in real time
between sensor nodes. The acquisition cost of such high bandwidth communications
systems is usually very significant. Network scalability is also affected by the amount of
allocated bandwidth. Hence to ensure that the sensor network operates and is scalable to
some reasonable degree, very high bandwidths must typically be allocated.

12
 In order to solve the bandwidth problem, the system must make use of
dynamically changing partial information from remote sensors by requesting sensor data
in an optimal manner so that the most useful sensor data is sent over the currently
available network bandwidth (no matter how small the available bandwidth is).

The time sequence which specifies the best sensor data to utilise is called the
optimal sensor schedule . Many currently implemented sensor scheduling algorithms for
distributed sensors employ ad hoc sensor scheduling techniques. The problem with such
approaches is the difficulty in quantifying system performance in multi-target or dynamic
sensor and bandwidth conditions. Therefore, there is a need to develop a well founded
analytic approach to the distributed sensor scheduling problem based on stochastic sensor
scheduling and control. The theory can be applied to the distributed multi-sensor
estimation problem where there are time-varying communication bandwidth constraints.
The underlying problem of stochastic sensor scheduling with system constraints,
however, presents a computational burden.

The significance of the sensor scheduling and optimization are

13
 Sensing accuracy: The utilization of a larger number and variety of sensor
nodes provides potential for greater accuracy in the information gathered as compared to
that obtained from a single sensor. The ability to effectively increase sensing resolution
without necessarily increasing network traffic will increase the reliability of the
information for the end user application.

Area coverage: A distributed wireless network incorporating sparse network

properties will enable the sensor network to span a greater geographical area without
adverse impact on the overall network cost.

Fault tolerance: Device redundancy and consequently information redundancy

can be utilized to ensure a level of fault tolerance in individual sensors.

Connectivity: Multiple sensor networks may be connected through sink nodes,

along with existing wired networks (eg. Internet). The clustering of networks enables
each individual network to focus on specific areas or events and share only relevant
information with other networks enhancing the overall knowledge base through
distributed sensing and information processing.

Minimal human interaction: The potential for self-organizing and self-

maintaining networks along with highly adaptive network topology significantly reduce
the need for further human interaction with a network other than the receipt of
information.

Operability in harsh environments: Robust sensor design, integrated with high

levels of fault tolerance and network reliability enable the deployment of sensor networks
in dangerous and hostile environments, allowing access to information previously
unattainable from such close proximity.

14
 Dynamic sensor scheduling: Dynamic reaction to network conditions and the
optimization of network performance through sensor scheduling. This may be achieved
by enabling the sensor nodes to modify communication requirements in response to
network conditions and events detected by the network, so that essential information is
given the highest priority.

Current sensor network applications include military sensing, air traffic

control, video surveillance, traffic surveillance industrial and manufacturing
automation, robotics, infrastructure monitoring and environment monitoring. Future
applications and capabilities may include the following :

Surveillance: A major current challenge for border surveillance is to deliver the

benefits associated with intelligent sensor networks using the existing low bandwidth
communications infrastructure. However, even very simple sensor networks have
proven to be very complex to properly control resulting in a number of ad hoc design
procedures yielding severely limited realizable benefits from the data fusion system.
The development of techniques for distributed data fusion within such sensor networks
is thus required.

Supply chain management: An important application arises in the well known

problem of supply chain management in a warehouse. Several tens of mobile Personal
Digital Assistants (sensors capable of transmitting images, text and voice) interact with
central sophisticated servers provide command and control solutions for smooth
delivery of products and maintenance of inventory. Sensor scheduling algorithms are
immediately applicable to this application.

Data fusion and tracking :

Data fusion is one the fundamental elements of modern tracking techniques,

utilising information from a variety of sources and combining the information in a way
that meets the desired application constraints and objectives.

15
 Data fusion is a framework describing the process of combining data originating
from different sources. The objective of data fusion is maximization of useful
information, such that the fused information provides a more detailed representation with
less uncertainty than that obtained from individual sources. While producing more
valuable information, the fusion process may also allow for a more efficient
representation of the data. Another by-product of information fusion may be the
observation of higher-order relationships between respective entities.

The selected method for performing the data combination will depend on the
original data format produced by the various sensor types. Data fusion is, in general,
conducted using one of the following frameworks:

• Pixel level fusion

• Feature level fusion

• High-level data fusion

Pixel level fusion describes the combination of multiple images into a single
image, where raw data is robustly and redundantly merged. Each location in the resulting
image is an algorithmic combination of the vector of measurements from each of the
sensors.

Feature level fusion refers to the extraction of features from each of the sensor
data. Registration of detected features is performed for regions of interest or image
segments containing more than one pixel. A detection/classification algorithm can then
be applied on the combined feature vector.

High-level data fusion or decision fusion occurs where sensor data, with or
without pre-processing, is combined with other data or a priori knowledge. Each sensor
makes an independent decision based on its own observations and passes these decisions
to a central fusion module where a global decision is made.

16
 Alternatively, in a decentralized multi-sensor system each node functions
performs data fusion based on local observations and the information communicated
from neighboring nodes.

The significance of the data fusion and tracking are

• Maximization of useful information

• More reliable information than possible from individual sources
• More efficient data and information representation
• Detection of higher-order relationships between different dedicated sensor types

Shown below (Figure 1) are some of the techniques applied to the various elements of the
data fusion process.

Practical applications of data fusion have necessarily been those areas in which the
required output of an analysis may not be measured directly.

This is particularly important such as:

17
 • Medical imaging
• Non-destructive testing
• Remote sensing applications such as target identification and tracking
• Condition monitoring for the detection of faults and degradation of machinery
• Landmine detection

INFORMATION PROCESSING

Information processing encompasses the data related processing that aims to

enhance and interpret the collected data and maximize the efficiency of the system,
through signal conditioning, data reduction, event detection and decision making. This
may involve a collection of filtering and other data manipulation techniques together with
advanced learning techniques for feature extraction and classification in order to provide
the most relevant data in an efficient representation to the communications interface.

The development of new techniques is required to process the large volumes of

information produced by sensor networks and adaptively implement the necessary
response.

This may encompass:

1. sensor scheduling
2. decision theory
3. feedback theory
4. state estimation
5. tailored supervisory control systems control systems
6. optimal sensor location
7. pattern recognition
8. data mining
9. network flow control
10. multi-resolution data transmission integrated with data fusion and reconstruction.

There are various applications for the information processing

18
 Surveillance and Monitoring Surveillance and monitoring is an application
which places an ever increasing demand on the advancement and development of new
information processing techniques.

Data analysis: The operation of surveillance and monitoring systems implies the
detection and tracking of an event or target. Based on the sensor type and resulting data
output, appropriate algorithms and processing techniques are required to extract this
information in real time, in order for other systems to be able to react in some way, either
by providing and alert to a sub-system or modifying the behaviour of the surveillance
system, such as tracking a given target following positive identification.

Data reduction: With the continual introduction of new sensor technologies and
the influx of information previously unavailable, considerable processing of the gathered
data is required to enable efficient dissemination of the data.

Data interpretation: In order to obtain meaningful interpretations of the large

volumes and various forms of data obtained from various sensor type and networks,
additional processing of the information is required.

Tracking , fusion and vision systems :

Neuro biology:

1. Moving target detection by insects - figure/ground discrimination (dragonflies,

hoverflies)
2. Motion coding / adaptation (at least another 10 years basic physiology remains in
this area even to complete present project directions): This is the source for
algorithms that feed following projects
3. Computer modeling of insect vision algorithms
4. Natural image coding/ natural time series analysis

VLSI Robotic sensor fusion:

1. Adaptive motion chips (based on both of the above): Collaboration with Tanner
Research Inc
2. Spatial imagers (low pixel count imagers for feedback control systems)

APPLICATIONS OF SENSOR NETWORKS

19
 Bushfire response using a low cost, typically dormant, distributed sensor
network early warning and localisation of bush fires can be achieved, hence saving life
and property, whilst reducing the cost of monitoring

Intelligent transportation low cost sensors build into roads and road signs can
assist to manage traffic flow and inform emergency services of traffic problems

Real-time health monitoring a nano-technology based bio-sensor network can

assist in monitoring an ageing population, and inform health care professionals in a
timely manner of potential health issues

Unmanned aerial vehicle surveillance swarms of low cost unmanned

autonomous and co-operating aerial vehicles could be deployed to conduct surveillance
and monitoring in remote or hostile environments

Water catchment and eco-system monitoring and management sensor

networks that keep track of water quality, salinity, turbity and biological contamination,
soil condition, plant stress and so on could be coordinated to assist environmentally
sustainable management of entire water catchment areas

Robotic landmine detection A sensor network for the detection and removal or
deactivation of landmines. A reliable sensor network will enable the safe removal of
landmines in former war zones, reducing the risk to those involved in the removal
process. The cost effectiveness of the network will aid in the its application throughout
third world nations where the after effects of war continue to take a toll on people living
in areas still containing live explosives. The utilization of advanced sensor technology to
detect explosives, will overcome difficulties in detection of un-encased landmines.

20
 CONCLUSION

The University of Melbourne has experience in managing a variety of ARC-

funded and other major programs (utilizing a comprehensive network of financial,
ethical, management and resource systems) over many years. The department where the
lead researcher comes from has a long history of strong track record in securing
competitive grants including several ARC special research centers, Cooperative Research
Centers, and State Government funded Initiatives, and collaborative research centers. An
effective and efficient financial management system is already in place to provide high
quality support for the management of research centers and networks

There has been a well established track record in running workshops, short
courses and conferences in areas related to this network. For example, the department
hosted the 6th International conference in Data Fusion 2003, will host the Asian control
conference in 2004, and will host the International conference in Intelligent Sensors in
December 2004. Several research assistants in the centre are well versed in web
authoring and on-line management of resources.

21
 REFERENCE LISTS

www.sensornetworks.net.au

www.research.rutgers.edu

www.jimpinto.com/sensornetworks.

22