Brendan Farrell’s HowLoud lets home-buyers know how much noise a location really generates
By Justin Chapman, Pasadena Weekly, 8/27/2015
Despite its reputation as a “sleepy” town, a new noise app developed by a former Caltech postdoctoral student and math instructor shows Pasadena has moderate to severe vehicle noise in all of the city’s zip codes.
HowLoud.net, the free online service developed by Brendan Farrell that allows users to generate noise reports about a specific address, was launched in Los Angeles and Orange counties this summer. The app features information on vehicle and air traffic, local businesses and other sources of noise pollution to generate a “noise rating” for that location, enabling people to research the amount of sound in the area before they commit to a new home or apartment building.
Farrell has a Ph.D. in applied mathematics from UC Davis, spent five years as a researcher in applied math and electrical engineering in Berlin and Munich, was a postdoctoral student in computing and mathematical sciences at Caltech and taught applied and computational mathematics there for three years before leaving to start HowLoud. He got the idea for the app when he and his wife were house hunting in Silver Lake last year.
“We were looking for a new place, and I assumed you could find this kind of information, but you couldn’t,” said Farrell. “You could find out everything else, but [noise information] wasn’t available.”
‘Striking’ Variability
The state of California requires cities to conduct noise studies, but HowLoud’s team rebuilds those studies from scratch “in a modern way,” said Farrell.
“[The studies are] not done for regular people and they’re done in kind of an antiquated way. On the tech side of it, we’re doing it in a modern way, and on the use side of it, we’re doing it with the intention to deliver just useful, simple stuff for people,” Farrell said. “So if you type in an address, like a hotel, or a street location, in five seconds you should be able to come up with something useful.”
Farrell said the app uses mathematical computation, data analysis and computer algorithms, combined with data on roads, airports and businesses, to create a complete map of noises around real estate.
According to the app, most of Pasadena’s zip codes have “severe” scores, with a couple of “moderate” scores, mostly due to vehicle noise. Some neighborhoods are quieter than others. Farrell said the variability within neighborhoods is “striking” and unpredictable.
For example, the Pasadena Weekly reported in 2009 on a group of Northwest Pasadena residents who were angry about the amount of noise generated by police helicopters. They said the copters flew directly over their homes at all hours of the day and night, constantly disrupting their children’s sleep and invading their privacy.
“It feels more like surveillance than crime prevention,” local activist Ricardo Costa said at the time. “I wouldn’t mind as much if they flew over Caltech just as much as they fly over my neighborhood.”
The Noise Element of the city’s General Plan, which “provides policy-level direction for the city to limit people’s exposure to noise,” acknowledges the contribution of helicopters to the city’s noise environment, as well as the complaints they generate. Elements of the General Plan were recently updated, though the Noise Element was last updated in 2002.
Federal officials have also weighed in on this issue. Following a bill introduced by Congressman Adam Schiff (D-Burbank) and US Sen. Dianne Feinstein (D-CA) in 2014, the Federal Aviation Administration launched a helicopter noise complaint system for LA County in April. Residents can file a noise complaint by calling (424) 348-HELI (4354) or by visiting heli-noise-la.com.
“Helicopter noise has disrupted the daily lives of thousands of Los Angeles residents for years, and allowing the public to report incidents to the FAA is a welcome and necessary step toward solving this problem,” Feinstein said in a statement. “While the reporting system is important, additional action by the FAA is needed — and overdue. The agency must work with pilots and the public to propose new flight patterns and practices. We will continue to press the case for rules that can reduce noise and protect privacy in Los Angeles communities.”
Farrell said helicopters are the biggest noise source in LA County that HowLoud does not yet cover in its noise map, but he added that they are working on it and will cover helicopters in time.
“We intend to report the average frequency of helicopter flights over a given location and a breakdown by time of day,” he said. “Of course, collecting helicopter information is much more difficult than vehicle information.”
Farrell also acknowledged that it is difficult to determine air traffic noise, noting that the Santa Monica Airport has been the toughest thing his company has dealt with so far.
“We treat all the airports according to the same standards, based on how much traffic they have and that sort of thing,” he said. “We’ve gotten a lot more people emailing saying, ‘Hey, you said airport traffic noise was light near my house, well it sure as hell is not light.’ But that airport has a relatively small number of flights compared to LAX; we have to balance all those things.”
The Sound Score Tool
A press release from the company states that they don’t use thousands of microphones to determine a location’s noise level. “Rather,” Farrell explained, “we build a model and determine the sound profile created by the sources, such as vehicle flow with certain speed and volume, a certain type of plane flying overhead, or a stadium with thousands of cheering fans. We then use physics to propagate the noise through the environment. The noise is attenuated as it travels, is reflected off obstacles and has its frequency profile changed. Our model incorporates all these effects and gives the noise level in decibels. The human ear, however, cares about more than just the average decibel level. Time matters: noise in the night is worse than noise during the day.”
Farrell confirmed that it would be too expensive to deploy several thousand sensors for actual readings to cover the vast majority of locations. However, he spends a lot of his time with “a handful of measuring and recording tools for different jobs, most importantly to test our results and to train my ear for the decibel scale. Our results currently cover several million homes and are set to expand quickly.”
HowLoud has received extensive media attention in recent weeks, with coverage by LA Weekly, SoCalTech.com, Caltech’s Engineering & Science magazine, International Business Times, Patch.com, the Atlantic CityLab, Gizmodo India, Engineering.com, Sioux City Journal, LAist, Curbed SF, the Eastsider LA blog and the Johnny, Etc. blog.
Following a successful Kickstarter fundraising campaign in July, in which the company raised $38,923, Farrell and his team of three engineers plan to expand beyond Southern California.
“We developed the technology so far, we did this region and we’re just improving how everything functions,” Farrell said. “Then we’re going to expand to much larger regions. It’s not going to be a county-by-county type of thing. We’re setting up infrastructure such that we can really quickly do much larger regions.”
HowLoud is also beginning to partner with real estate companies and other interested investors by allowing them to embed the HowLoud display directly on their Web site, similar to the way a weather report, Google Maps, or many other applications can be embedded.
Farrell said the app is not meant to tell people whether they should purchase a certain house. Rather, it is one more tool for consumers to reference when researching new neighborhoods.
“It’s not always about, ‘Hey, should I buy that house or not? Well, let me check HowLoud and see what they say about the site.’ It’s more like, ‘Should I drive across town to visit this place?’ We’re not trying to tell you whether you should buy a house or not, but maybe whether or not you should continue reading about that listing or whether you should drive across town to see it. If we say, ‘It’s terribly loud,’ and that matters to you, well you save yourself an hour to drive over and see it.”
So just how accurate is HowLoud? Farrell said it’s somewhat subjective.
“From one house to the house next door, I think it’s very accurate,” Farrell said. “Accuracy is kind of tough in that what we do is a little bit like ‘How good is that movie?’ or ‘How good is that restaurant?’ There’s only one component that’s a scientific quantity, the vehicle decibel estimate. Otherwise, the sound score is a tool that we’ve come up with that incorporates different factors and weighs them accordingly. The goal is to be useful. We expect people to check five addresses, see what we say, and if that agrees with your knowledge of those locations, then you’ll trust us.”