The latest news from academia, regulators
research labs and other things of interest
Posted: Apr 04, 2011
Body sensors are joining the future internet
(Nanowerk News) Modern man is surrounded by a multitude of sensors. Today's sensors are simple and made for specific purposes, like measuring temperature, balance, build-up of smoke, or tyre pressure. The common trait of the sensors is that they are all embedded in a closed system.
Now researchers at the Department of Informatics, University of Oslo, are developing sensors for the future. Their size decreases. They are more robust than today's sensors, communicate wirelessly, and even reduce energy consumption to a minimum. But more important: The researchers are connecting all sensors to the Future Internet.
"We are developing a completely new paradigm within computer science, where we look at how to effectively exploit networked sensors. Computer programs can search for suitable sensors and employ them without up front knowledge about which sensors are actually available. The programs are also indifferent to which networks are present," explains Professor Thomas Plagemann at the Department of Informatics in University of Oslo in Norway.
He is head of the new SET (Smart Environment Technology) initiative. This is an interdisciplinary cooperation between as disparate disciplines as nanoelectronics, digital signal processing and distributed multimedia systems.
In order to meet the objectives the researchers must take into account a wide range of subjects, like wireless network technology and radio waves, ultra sound and infrared signals, signal processing, middleware, optimally collecting and aggregating data into databases, and security. They must also define formal concepts and new methods and standards.
By connecting the sensors to the Internet the computer can control the sensor information and find the optimal way of making use of the sensors.
"The new technology has immense possibilities. Examples include monitoring ocean currents, global warming, pollution, energy consumption, traffic, pollen dispersal, animal migration, and others. The future sensor network is also well suited for home care," says Thomas Plagemann to the research-magazine Apollon at University of Oslo.
Under Skin Sensors for Diabetes Monitoring
In cooperation with The Intervention Centre at Oslo University Hospital, Vestfold University College, and LifeCare in Bergen, the researchers are already working on the development of a brand new type of sensor that can be placed under the skin for measuring the blood glucose level.
"This nano instrument is only three times seven millimeters wide and consequently small enough to be injected underneath the skin. No operation is needed. A nano membrane moves when the blood sugar concentration is irregular," says Professor Tor Sverre Lande at the Department of Informatics.
He is an expert on sensors and networks and has previously developed an artificial ear, the reknown cochlear implant, that provides a sense of sound to deaf persons.
Philipp Häfliger (right) and Tor Sverre Lande have developed a small instrument that measures blood sugar concentration underneath the skin. When sensors are connected to the Internet, researchers can monitor all diabetes patients in Oslo simultaneously and investigate the relationship between insulin level and physical activity. The society can also use sensor networks for automating elderly care and monitoring nuclear disasters.
It is difficult to make body implant sensors that function over a long time span. All sensors need energy.
"A battery is too large. And we do not want to replace batteries within the body. Preferably batteries should be avoided. They can leak chemicals," Tor Sverre Lande explains.
The researchers have consequently developed wireless energy transfer between the glucose sensor and a bracelet carrying a battery.
Such wireless energy transfer is not more advanced than that of a standard high voltage current transformer. A magnetic field between two coils of wire causes voltage changes.
Exactly the same principle is applied to the diabetes patients.
"We have coils both underneath and outside the skin. Then the energy is transferred by electromagnetic waves."
During wireless energy transfer most of the energy is lost. Only two percent of the energy is caught by the skin-deep coil. The reason is that the antenna in the nano instrument is very small.
"It is therefore important to develop a glucose meter that uses as little energy as possible," explains Tor Sverre Lande.
The researchers also want to connect the glucose body implant to the Future Internet. By combining different technologies and measuring the level of insulin and how much the patients move, researchers can study the relationship between insulin level and physical activity.
It is easy to monitor patient movement using the built-in accelerometer of a modern mobile phone.
"By collecting data from all diabetes patients in Oslo we can obtain a detailed picture of how many diabetes patients are exercising. This can improve diabetes treatment. We also want the body sensor to talk with the mobile phone. Then the mobile phone can give a ring when the insulin level is too low, or say: "Take a walk"," envisions Plagemann.
Automated Elderly Care
Even elderly people and persons with dementia can benefit from networked sensors. The researchers at the Department of Informatics are investigating how sensors can be used in automated home care. The system among other things makes an alert if someone drops to the floor, or a dementia patient slips out of the apartment or forgets to turn off the stove.
"With such a system more elderly people can remain in their homes. This increases their well-being," Thomas Plagemann believes.
Different sensors can measure if a person lies on the floor. Some elderly will have camera surveillance. Others have an accelerometer in their mobile phone that measures if the patient moves.
"Our challenge is the large diversity of sensors. And the coverage area of the sensors is affected by the shape of the room."
Describes Diverging Behaviour
Plagemann's research group is currently making a high-level description of diverging behaviour.
"The computer program shall automatically detect sensors that offer the property "Person falls"."
The system must integrate many different technologies as it is not certain that all sensors in the apartment make correct observations. A camera sensor can trigger a false alarm if it registers a person having a sun bath outside the window.
If a motion sensor reports that a person has tumbled, it might happen that the person simply overdid some arm-swinging. And the demented person may not have run away. Maybe he is just visiting friends. The number of potential sources of error is large.
The system must therefore, completely automated, find all the sensors in the apartment, combine their information, and make a likely picture of what has happened.
The Future Internet is an important ingredience. In addition it is important with a swift interpretation of the data. It is of no use if the alarm goes off fifteen minutes after a heart failure.
Checking Nuclear Disasters
When sensors are connected to a network, they can also be used for supervising contaminated areas. Ground personnel had to move into the disaster area to make measurements of the radiation when the world woke up to the Chernobyl nuclear disaster in 1986.
At the next nuclear disaster sensors can be dropped from aircrafts to check radiation in the contaminated area.
"If the sensors can talk to each other, one can collect information from the radioactive area by connecting to the outermost sensor. Then one can avoid sending humans into the contaminated area," explains Tor Sverre Lande.
The main limitation of the sensors is the energy supply.
Sensors deployed in the nature need to function for a long time without the need for a battery change. Moreover it is important that the sensors can communicate and forward information even if the area does not offer network access. To transfer signals between two sensors the distance must not be too large.
The explanation is simple: The energy needed for sending signals over twice the distance, quadruples. That is: If it costs ten energy units to send a signal ten meters, then it costs 100 energy units to send the signal 20 meters. It is therefore better with many short distances between the sensors than one long distance.
If one wants a database in the sensor, it is important to be able to use it with as little energy consumption as possible. When the database is designed, it is therefore necessary with a detailed understanding of the sensor.
Computer scientists are structured persons. They are used to organizing everything in many layers, like placing hardware at the bottom. Then comes network and communication layers. And at the top they place databases and applications. Such a way of organizing matters requires internal communication between the layers. Consequently extra computation and energy are needed.
"To minimize the sensor energy used for communication between the database and the hardware, it has been necessary to diverge from this structure," explains Thomas Plagemann.
Sensor networks can also be used to monitor global warming in polar areas or areas with a danger of landslides or avalanches.
Typically the mobile phone coverage is bad in such areas. The sensors can communicate and forward the information to chance passers-by that afterwards carry the information with them on their mobile phone. Then the information can be passed on when within a coverage area.
Weather Forecast using Cars
Modern cars are loaded with sensors. The researchers want to connect these sensors to the Future Internet.
An example of a sensor that is exciting for the society, is the rain sensor of modern cars.
In the future the rain sensors can communicate. Then the meteorologists can collect information from the sensors and update the weather chart. Instead of each car sending data to a shared weather database, it is more practical if the system collects information from a handful of cars while one's at it.
The railway administration can also make good use of a sensor network. They can use sensor networks to make alerts at landslides, by placing sensors with radars along the tracks. The system can also give notice if the rails need snow removal.
"Only the imagination sets a limit to what a sensor network can be used for," says Plagemann.
Sensors can also be placed in the concrete mix in new buildings. Then the sensors can inform if the buildings have been damaged after an earthquake.
By analyzing the background noise of all telephone calls it is possible to make a noise map of Oslo.
One can also use sensor networks to investigate possible relationships between the number of people and high humidity in a building.
Advanced Control Systems
Professor Oddvar Søråsen, head of the Nanoelectronics research group at the Department of Informatics in University of Oslo, is interested in how computers that interpret sensor data can control actuators.
In contrast to a sensor, that only observes, an actuator influences its surroundings. Actuators are small switches, valves, or robots that can perform simple tasks like regulating nerve signals or the fluid stream in a small chip. Actuators can also be used for intelligent air regulation, automatic control of livestock during the breeding season, or smart treatment of sewage and wastewater at heavy rainfalls and floodings.
To make new actuators the researchers want to combine micromechanics and microelectronics. This is a new field of research.
Today the micromechanics and the microelectronics are provided by two different chips. If we can produce both on one chip, we save communication between two chips. Then the power requirement is reduced.
When researchers design the nanopattern of a silicon chip, they can reserve parts of the chip for mechanical motion.
About sensor networks
In the networks of the future, sensors are connected to the Internet. This is called a sensor network.
The sensors will also communicate between themselves.
A sensor network offers immense possibilities and can among other things be used for elderly care, patient research, concrete monitoring, surveying contaminated areas, and traffic monitoring. Only the imagination sets the limit.