Skip to Content

Where Sensors Make Sense

Siemens aims to turn your thermostat into a “comfortstat” – and create a viable market for wireless sensor networks.
December 15, 2005

The idea of tiny, ubiquitous computers monitoring us and our environments from every nook and cranny might alarm a few civil libertarians – but this is exactly the concept driving researchers who are trying to perfect networks of smart, wireless sensors.

They envision sensors sprinkled across a battlefield to warn of an enemy advance, or attached to pill bottles to alert caregivers to when an elderly patient takes (or doesn’t take) his or her medication. They imagine faulty equipment in manufacturing plants that reports its own failures. In short, they see a pervasive grid of smart sensors that monitor, analyze, and network the bits and bytes of life.

For years, these networks have remained largely in the prototype stage, not quite ready to hit the market. “We know that it is a good concept; we’ve demonstrated its feasibility,” says Osman Ahmed, senior principal engineer at Siemens Building Technologies in Buffalo Grove, IL. The problem, Ahmed explains, is that no major company has yet found a compelling commercial market for sensor networks.

Research groups at Intel and the University of California’s Berkeley and Los Angeles campuses continue to experiment with research prototypes. Small companies, such as Sensoria in San Diego and Ember in Boston, have pushed the technology forward, but have yet to break through with a killer app.

Siemens, however, has developed a new type of sensor that can detect multiple types of environmental changes and is mass-producible. And the company is studying which sensor network applications will appeal to customers first.

In most sensor networks, each node includes an individual sensor that detects one of several parameters such as temperature, humidity, vibration, or the presence and concentration of a chemical. The sensor is coupled to a microprocessor that stores and processes information and a radio that transmits data. These nodes are dispersed throughout an environment, such as city sewer lines or building spaces, and communicate via a “mesh network” in which information hops from node to node, en route to a centralized hub.

What makes Siemens’ devices a breakthrough is that six different sensors can be fabricated on one chip, layered on top of each other, “like a sandwich,” says Ahmed. Each layer is made of a material that’s specialized for reading a different environmental element. For instance, when carbon monoxide molecules land on the layer that’s sensitive to it, the electrical properties of that layer change with the concentration of the gas. This electrical information is transmitted to the microprocessor, then relayed to the network’s central hub.

By streamlining the fabrication process and putting six sensors on one chip, Siemens has reduced production costs to a sixth of the former cost, says Ahmed.

Siemens has also reduced the sensors’ power consumption by using power-scavenging techniques. Instead of relying solely on a lithium-ion battery, the sensor is designed to acquire energy such as ambient light from its surroundings. The sensors also save power by going into “sleep” mode when the environment is static and information doesn’t need to be transmitted, since sending information is a power-hungry operation.

With the production costs of sensors and their demand for power decreasing, Ahmed predicts that the one of the first applications for them will be “comfortstat,” as he likes to call it. “It would give you an index of comfort, not just temperature,” he says. Ahmed envisions a sensor in every room of a home, collecting information about humidity, airflow, and levels of carbon dioxide, carbon monoxide, and volatile organic compounds, such as those in paint or aerosol sprays.

These sensors would send environmental information back to the comfortstat, which would display the particular conditions in each room and alert people about any hazards. This application would be especially useful in large buildings, where the sensors could identify, for instance, pockets of heat, saving energy and money, Ahmed says.

Another early application being tested by researchers at the University of Florida in Gainesville, with sponsorship from Siemens, is using sensor networks to monitor small animal cages in research laboratories. While unglamorous, this application could make the large mouse colonies needed for genetic and pharmacological research easier to maintain and cheaper to operate.

Ahmed explains that his sensors would track ammonia levels in the cages – high ammonia levels in urine can mean an infection – as well as temperature and water levels (overactive mice sometimes jam the nozzles on their water bottles). These lab environments need to be scrupulously monitored to ensure that the animals are responding to certain treatments, not unintended environmental elements. Each cage’s information can also be displayed so that care takers can find and fix a problem quickly.

Ahmed estimates that the comfortstat and mouse-cage applications are roughly two years from commercialization. Meanwhile, Siemens is beginning market research to determine how best to sell its sensors.

“Siemens’ application is an excellent example of exploiting available capabilities” of sensor networks, says Deborah Estrin, director of UCLA’s Center for Embedded Networked Sensing.

But Estrin says the challenge with such multisensor networks, especially as the sensors become more versatile, is sorting through all the data collected at each sensor, and ultimately at the hub.

Ahmed agrees that people shouldn’t have to deal with the raw data from large networks of sensors. “The whole idea is to not really throw out a bunch of sensors, but to actually create value,” he says. To accomplish that, his team is working on data-mining algorithms that can pick out useful information, as well as machine learning algorithms that can predict which information will be useful based on prior data.

It may be a while longer before the typical house is equipped with a comfortstat. Instead, businesses with large buildings or manufacturing plants, where the high initial cost of installing a large number of wireless environmental sensors is more justifiable, will most likely adopt the technology first, says Ralph Kling, director of Intel’s Sensor Network Operation. “You reach a higher volume…and it becomes more attractive to users,” he says.

For now, though, a successful business plan is still the missing element in the world of wireless sensors. “I kind of learned from my own experience that if you want to commercialize technology, the technology isn’t the barrier,” says Ahmed. “It’s how we market it.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.