快猫短视频

Anthony Steed wanted to know how polluted the air was on his daily cycle commute. So in April 2005 he packed a backpack with a carbon monoxide sensor, GPS receiver and a data recorder, and set off to his office in the computer science department at University College London.

After wearing the backpack for two months along the same route he and his colleague Richard Milton analysed the data. They found that just as Steed passed over a particular railway bridge, the sensor registered a large spike in the level of carbon monoxide.

Like most cities, London has only a few dozen permanent, fixed pollution sensors 鈥� too few to provide much local detail of the pollution city-dwellers are exposed to each day, like the high levels of carbon monoxide at the railway bridge. Milton and Steed鈥檚 research aims to fill in the gaps. It began in 2003 when volunteers wearing the backpacks sampled the pollution along a stretch of London鈥檚 busy Marylebone Road. Now, inspired by Milton and Steed, researchers around the world are hoping to revolutionise pollution monitoring. By connecting cheap, commercially available sensors to wireless devices and teaching volunteers how to use them, they want to create simple sensing networks that blanket entire cities.

Networks such as these could tell people what pollutants they are exposed to over the course of the day and allow them to use this information to minimise any respiratory problems they might suffer, says Rufus Edwards, who investigates air pollution exposure at the University of California, Irvine. Doctors treating patients with breathing problems would know what they have been inhaling, while asthma sufferers could avoid danger areas when pollution concentrations are high. Harnessing these technologies could change the way health studies are done, Edwards says.

鈥淗arnessing these technologies can change the way we do health studies鈥�

His work is part of a broader move to more accurately measure people鈥檚 exposure to toxic chemicals. Last week the Californian state assembly passed a bill for a state-wide bio-monitoring programme to test volunteers for exposure to synthetic chemicals. The bill has yet to be signed by governor Arnold Schwarzenegger, who vetoed a similar bill last year.

To measure pollutants in the air, Edwards has worked with Kirk Smith of the University of California, Berkeley, to turn a commercial smoke detector into a low-cost sensor that can detect ultra-fine particles. The smoke detector contains an LED and a light detector. When particles in the air disrupt the beam, the sensor detects the scattered light, and the amount by which the particles scatter the light provides a measure of their concentration.

The cheap sensor is allowing the researchers to carry out extensive public health studies at a fraction of the previous cost. 鈥淲hat was once done using a $4000 to $6000 device we鈥檝e been able to do for $150 in parts,鈥� says Edwards.

The Californian team is now taking the sensors further afield to measure levels of ultra-fine airborne particles in homes in Guatemala, India, Nepal and China. In these countries, the use of wood or crop-burning stoves means indoor concentrations of particles of ferric iron, manganese and carbon compounds can reach thousands of micrograms per cubic metre. This compares with average levels of 20 to 40 micrograms per cubic metre in developed countries.

Edwards has also been working with artist and technology researcher Beatriz da Costa, also at the University of California, Irvine, on a project called the PigeonBlog. They let loose a flock of pigeons, each fitted with a GPS receiver, air pollution sensors and a basic cellphone, in the skies above California (快猫短视频, 4 February, p 29). Last month the birds flew over the cities of San Jose and Irvine, and another release is planned over Newport Beach on 18 September. The devices collect data on pollutant levels and beam it via text messages to appear on a web page in the form of an interactive map.

Next week da Costa will be launching another project, called AIR (Area鈥檚 Immediate Reading), in which 10 hand-held sensors will be used to map pollution in lower Manhattan. The sensors will continuously measure carbon monoxide and nitrogen oxides, and the data will be transmitted to a central server to create a real-time map. As well as sensors and a GPS receiver, the hand-held devices have a database of known pollution sources, such as power plants and factories, allowing the volunteers to see how far they are from these sites and check the effect on local air quality.

People will be able to use the devices without special training, says Brooke Singer, a digital media artist and co-creator of AIR. 鈥淲e were going for simplicity and ease of use. This will help ordinary people participate in the conversation about air quality issues 鈥� a conversation they don鈥檛 usually have access to.鈥�

Even more ambitiously, Shannon Spanhake, a new-media artist, is hoping to create a network of hundreds of volunteers to monitor pollution in San Francisco. Spanhake has developed a cellphone-sized device that contains sensors capable of detecting carbon monoxide, nitrogen oxides, sulphur dioxide and ozone. Each sensor consists of a metal oxide chip across which a voltage is applied. Changes in gas concentrations perturb the flow of electricity across the chip, which is calibrated to a specific pollutant, so measuring the electric current reveals the current gas concentration. While monitoring air quality, the devices will detect nearby Wi-Fi hotspots and upload information to a server to map the data.

Spanhake will launch the project in October through San Francisco State University, where she will teach students to build the devices themselves. Once enough volunteers have signed up, she hopes to use the city鈥檚 planned municipal Wi-Fi infrastructure to create a pollution monitoring network.

To carry out a wide-scale public health study investigating personal pollution exposure, such networks will eventually need to expand to include thousands of volunteers. Organising such large numbers of people will be a challenge, says Michael Jerrett, who investigates the relationship between the environment and health at the University of Southern California in Los Angeles. 鈥淭he devices have to be charged each night and people have to remember to wear them every day. Or people may just think 鈥楬ey, this is a neat toy,鈥� decide to keep them, and you鈥檒l never hear from them again,鈥� he says. This is likely to become less of a problem as the price of sensors falls, and low-cost wireless devices such as da Costa鈥檚 and Spanhake鈥檚 are developed.

Another difficulty is maintaining the quality of the data. Most air-quality researchers up to now have used data from a few calibrated, high-precision sensing stations spread out across urban areas. Now, thanks to their low-cost sensors, Smith and Edwards are finding themselves flooded with information to be analysed. 鈥淚n the past, we always tried to maximise data output,鈥� says Edwards. 鈥淣ow we could go out and stick 200 devices out there, but our ability to process the information is becoming the limiting factor.鈥�

To this end, the researchers are attempting to improve their data-processing software, and they believe that as their experience grows, so will their ability to handle large amounts of information. 鈥淭hese technologies are still in the development phase, but already they鈥檙e proving immensely useful,鈥� Smith says.

鈥淥ne can easily envision city streets lined with sensors, wirelessly networked and sending data streams down to a central server,鈥� he says. 鈥淭hey would not just be for air pollution either 鈥� we could have sensors for water, soil and other parameters. It could open up a whole new realm of environmental science.鈥�