While things like rain or hills can keep people from cycling, a major impediment is the risk of getting hit by a car. It’s hard to identify the safest routes to ride, especially for beginner cyclists, and a key way to flag dicey streets involves time and injury: waiting until cars have hit several cyclists at a given location.

A 91̽��-led team has developed a system, called ProxiCycle, that logs when a passing car comes too close to a cyclist (within four feet). A small, inexpensive sensor plugs into bicycle handlebars and tracks the passes, sending them to the rider’s phone. The team tested the system for two months with 15 cyclists in Seattle and found a significant correlation between the locations of close passes and other indicators of poor safety, such as collisions. Deployed at scale, the system could support mapping or navigating cyclists on safer bike routes through cities.

“Experienced cyclists have this mental map of which streets are safe and which are unsafe, and I wanted to find a simple way to pass that knowledge down to novice cyclists,” said lead author , a 91̽��doctoral student in the Paul G. Allen School of Computer Science & Engineering. “Cycling is really good for your health and for the environment. Getting more people biking more often is how we reap those rewards and increase safety in numbers for cyclists on the roads.”

Apr. 29 at the ACM CHI Conference on Human Factors in Computing Systems in Yokohama, Japan.

To start, researchers surveyed 389 people in Seattle. Respondents of all cycling experience levels ranked the threat of cars as the factor which most discouraged them from cycling, and said they’d be very likely to use a map that helps navigate for safety. But a key factor preventing this is limited data on road safety.

The team then built a small sensor system that plugs into a bike’s left handlebar. The system, which costs less than $25 to build, consists of a 3D printed plastic casing that houses a pair of sensors and a Bluetooth antenna. The antenna transmits data to the rider’s phone, where the team’s algorithm susses out what’s a passing car rather than a person, or another cyclist, or a tree.

The team validated the system both by testing it in a parking lot, with a car passing at different distances, and with seven cyclists riding through Seattle with GoPro cameras on their handlebars. Researchers watched the footage from these rides and compared this to the sensor output.

The team then recruited 15 cyclists through the newsletter of Seattle Neighborhood Greenways, a local advocacy group. Each got a ProxiCycle sensor, a custom Android application and instructions. The cyclists took 240 bike rides over two months and recorded 2,050 close passes. Researchers then compared the locations of close passes with riders’ perceived safety at different locations in the city — which they measured by showing cyclists images of locations and having them rate how safe they felt at those locations (referred to as “perceived safety”) — and with the locations of known automobile-to-bike collisions in the last five years.

The team found a significant correlation between close passes and both other indicators of cycling risk. They also found that this measure of close passes was a better indicator of actual safety than the surveyed perceived safety, which is the current standard used by policymakers to study safety when collisions aren’t enough.

In the future, researchers hope to scale the study up and potentially account for other risk factors, such as cyclists being hit by opening car doors, and deploy ProxiCylce in other cities. With enough data, existing bike mapping apps, such as Google Maps or Strava, might include safer route suggestions for cyclists.

Some of those routes involve only minor adjustments.

“One study participant, who was living down by Seattle Center, was biking down Mercer all the time,” Breda said. “It’s this busy, multi-lane road. But just before the study, they found out that there’s a great bike lane on a quieter street, just one block north.”

, an applied science lead at Gridware, and , a professor at Georgia Institute of Technology, are also co-authors on this paper. , a 91̽��professor in the Allen School, is the senior author.

For more information, contact Breda at joebreda@cs.washington.edu.

Smart accessories are increasingly common. Rings and watches track vitals, while Ray-Bans now . Wearable tech has even broached . Yet certain accessories have yet to get the smart touch.

91̽�� researchers introduced the Thermal Earring, a wireless wearable that continuously monitors a user’s earlobe temperature. In a study of six users, the earring outperformed a smartwatch at sensing skin temperature during periods of rest. It also showed promise for monitoring signs of stress, eating, exercise and ovulation.

The smart earring prototype is about the size and weight of a small paperclip and has a 28-day battery life. A magnetic clip attaches one temperature sensor to a wearer’s ear, while another sensor dangles about an inch below it for estimating room temperature. The earring can be personalized with fashion designs made of resin (in the shape of a flower, for example) or with a gemstone, without negatively affecting its accuracy.

Researchers Jan. 12 in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies. The device is not currently commercially available.

“I wear a smartwatch to track my personal health, but I’ve found that a lot of people think smartwatches are unfashionable or bulky and uncomfortable,” said co-lead author , a 91̽��doctoral student in the Paul G. Allen School of Computer Science & Engineering. “I also like to wear earrings, so we started thinking about what unique things we can get from the earlobe. We found that sensing the skin temperature on the lobe, instead of a hand or wrist, was much more accurate. It also gave us the option to have part of the sensor dangle to separate ambient room temperature from skin temperature.”

Creating a wearable small enough to pass as an earring, yet robust enough that users would have to charge it only every few days, presented an engineering challenge.

“It’s a tricky balance,” said co-lead author , who was a 91̽��masters student in the electrical and computer engineering department when doing the research and is now at the University of California San Diego. “Typically, if you want power to last longer, you should have a bigger battery. But then you sacrifice size. Making it wireless also demands more energy.”

The team made the earring’s power consumption as efficient as possible, while also making space for a Bluetooth chip, a battery, two temperature sensors and an antenna. Instead of pairing it with a device, which uses more power, the earring uses Bluetooth advertising mode — the transmissions a device broadcasts to show it can be paired. After reading and sending the temperature, it goes into deep sleep to save power.

Because continuous earlobe temperature has not been studied widely, the team also explored potential applications to guide future research. In five patients with fevers, the average earlobe temperature rose 10.62 degrees Fahrenheit (5.92 degrees Celsius) compared with the temperatures of 20 healthy patients, suggesting the earring’s potential for continuous fever monitoring.

“In medicine we often monitor fevers to assess response to therapy — to see, for instance, if an antibiotic is working on an infection,” said co-author , a clinical instructor at the Department of Emergency Medicine in the 91̽��School of Medicine. “Longer term monitoring is a way to increase sensitivity of capturing fevers, since they can rise and fall throughout the day.”

While core body temperature generally stays relatively constant outside of fever, earlobe temperature varies more, presenting several novel uses for the Thermal Earring. In small proof-of-concept tests, the earring detected temperature variations correlated with eating, exercising and experiencing stress. When tested on six users at rest, the earring’s reading varied by 0.58 F (0.32 C) on average, placing it within the range of 0.28 C to 0.56 C necessary for ovulation and period tracking; a smartwatch varied by 0.72 C.

“Current wearables like Apple Watch and Fitbit have temperature sensors, but they provide only an average temperature for the day, and their temperature readings from wrists and hands are too noisy to track ovulation,” Xue said. “So we wanted to explore unique applications for the earring, especially applications that might be attractive to women and anyone who cares about fashion.”

While researchers found several promising potential applications for the Thermal Earring, their findings were preliminary, since the focus was on the range of potential uses. They need more data to train their models for each use case and more thorough testing before the device might be used by the public. For future iterations of the device, Xue is working to integrate heart rate and activity monitoring. She’s also interested in potentially powering the device from solar or kinetic energy from the earring swaying.

“Eventually, I want to develop a jewelry set for health monitoring,” Xue said. “The earrings would sense activity and health metrics such as temperature and heart rate, while a necklace might serve as an electrocardiogram monitor for more effective heart health data.”

, a doctoral student in the Allen School, was a co-author on the paper. , a professor in the Allen School, and , a professor in the Allen School and the electrical and computer engineering department, were co-senior authors. This research was funded by the Washington Research Foundation and the .

For more information, contact Xue at qxue2@cs.washington.edu and Liu at yul276@ucsd.edu.

For questions specifically for Dr. Mastafa Springston, please contact Susan Gregg at sghanson@uw.edu.

If you’ve ever thought you may be running a temperature yet couldn’t find a thermometer, you aren’t alone. A fever is the and an early sign of many other viral infections. For quick diagnoses and to prevent viral spread, a temperature check can be crucial. Yet accurate at-home thermometers aren’t commonplace, despite .

There are a few potential reasons for that. The devices can range from $15 to $300, and many people need them only a few times a year. In times of sudden demand — such as the early days of the COVID-19 pandemic — thermometers can sell out. Many people, particularly those in under-resourced areas, can end up without a vital medical device when they need it most.

To address this issue, a team led by researchers at the 91̽�� has created an app called FeverPhone, which transforms smartphones into thermometers without adding new hardware. Instead, it uses the phone’s touchscreen and repurposes the existing battery temperature sensors to gather data that a machine learning model uses to estimate people’s core body temperatures. When the researchers tested FeverPhone on 37 patients in an emergency department, the app estimated core body temperatures with accuracy comparable to some consumer thermometers. The team March 28 in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies.

“In undergrad, I was doing research in a lab where we wanted to show that you could use the temperature sensor in a smartphone to measure air temperature,” said lead author , a 91̽��doctoral student in the Paul G. Allen School of Computer Science & Engineering. “When I came to the UW, my adviser and I wondered how we could apply a similar technique for health. We decided to measure fever in an accessible way. The primary concern with temperature isn’t that it’s a difficult signal to measure; it’s just that people don’t have thermometers.”

The app is the first to use existing phone sensors to estimate whether people have fevers. It needs more training data to be widely used, Breda said, but for doctors, the potential of such technology is exciting.

“People come to the ER all the time saying, ‘I think I was running a fever.’ And that’s very different than saying ‘I was running a fever,’” said , a co-author on the study and a 91̽��clinical instructor at the Department of Emergency Medicine in the 91̽��School of Medicine. “In a wave of influenza, for instance, people running to the ER can take five days, or even a week sometimes. So if people were to share fever results with public health agencies through the app, similar to how we signed up for COVID exposure warnings, this earlier sign could help us intervene much sooner.”

Clinical-grade thermometers use tiny sensors known as thermistors to estimate body temperature. Off-the-shelf smartphones also happen to contain thermistors; they’re mostly used to monitor the temperature of the battery. But the 91̽��researchers realized they could use these sensors to track heat transfer between a person and a phone. The phone touchscreen could sense skin-to-phone contact, and the thermistors could gauge the air temperature and the rise in heat when the phone touched a body.

To test this idea, the team started by gathering data in a lab. To simulate a warm forehead, the researchers heated a plastic bag of water with a sous-vide machine and pressed phone screens against the bag. To account for variations in circumstances, such as different people using different phones, the researchers tested three phone models. They also added accessories such as a screen protector and a case and changed the pressure on the phone.

The researchers used the data from different test cases to train a machine learning model that used the complex interactions to estimate body temperature. Since the sensors are supposed to gauge the phone’s battery heat, the app tracks how quickly the phone heats up and then uses the touchscreen data to account for how much of that comes from a person touching it. As they added more test cases, the researchers were able to calibrate the model to account for the variations in things such as phone accessories.

Then the team was ready to test the app on people. The researchers took FeverPhone to the 91̽��School of Medicine’s Emergency Department for a clinical trial where they compared its temperature estimates against an oral thermometer reading. They recruited 37 participants, 16 of whom had at least a mild fever.

To use FeverPhone, the participants held the phones like point-and-shoot cameras — with forefingers and thumbs touching the corner edges to reduce heat from the hands being sensed (some had the researcher hold the phone for them). Then participants pressed the touchscreen against their foreheads for about 90 seconds, which the researchers found to be the ideal time to sense body heat transferring to the phone.

Overall, FeverPhone estimated patient core body temperatures with an average error of about 0.41 degrees Fahrenheit (0.23 degrees Celsius), which is in the clinically acceptable range of 0.5 C.

The researchers have highlighted a few areas for further investigation. The study didn’t include participants with severe fevers above 101.5 F (38.6 C), because these temperatures are easy to diagnose and because sweaty skin tends to confound other skin-contact thermometers, according to the team. Also, FeverPhone was tested on only three phone models. Training it to run on other smartphones, as well as devices such as smartwatches, would increase its potential for public health applications, the team said.

“We started with smartphones since they’re ubiquitous and easy to get data from,” Breda said. “I am already working on seeing if we can get a similar signal with a smartwatch. What’s nice, because watches are much smaller, is their temperature will change more quickly. So you could imagine having a user put a Fitbit to their forehead and measure in 10 seconds whether they have a fever or not.”

, a 91̽��professor in the Allen School and the electrical and computer engineering department, was a senior author on the paper, and , an assistant professor in the University of Toronto’s computer science department, was a co-author. This research was supported by the 91̽�� Gift Fund.

For more information, contact Breda at joebreda@cs.washington.edu. He’ll be traveling for research starting June 23; his availability for interviews will be limited after that.

For questions specifically for Dr. Mastafa Springston, please contact Susan Gregg at sghanson@uw.edu.