Tag Archives: Sustainability

A tale of a whale song

Like us, whales sing. But unlike us, their songs can travel hundreds of miles underwater. Those songs potentially help them find a partner, communicate and migrate around the world. But what if we could use these songs and machine learning to better protect them?

Despite decades of being protected against whaling, 15 species of whales are still listed under the Endangered Species Act. Even species that are successfully recovering—such as humpback whales—suffer from threats like entanglement in fishing gear and collisions with vessels, which are among the leading causes of non-natural deaths for whales.

To better protect those animals, the first step is to know where they are and when, so that we can mitigate the risks they face—whether that's putting the right marine protected areas in place or giving warnings to vessels. Since most whales and dolphins spend very little time at the surface of the water, visually finding and counting them is very difficult. This is why NOAA’s Pacific Islands Fisheries Science Center, responsible for monitoring populations of whales and other marine mammals in U.S. Pacific waters, relies instead on listening using underwater audio recorders.

NOAA has been using High-frequency Acoustic Recording Packages (HARPs) to record underwater audio at 12 different sites in the Pacific Ocean, some starting as early as 2005. They have accumulated over 170,000 hours of underwater audio recordings. It would take over 19 years for someone to listen to all of it, working 24 hours a day!


Crew members deploy a high-frequency acoustic recording package (HARP) to detect cetacean sounds underwater (Photo credit: NOAA Fisheries).

To help tackle this problem, we teamed up with NOAA to train a deep neural network that automatically identifies which whale species are calling in these very long underwater recordings, starting with humpback whales. The effort fits into our AI for Social Good program, applying the latest in machine learning to the world’s biggest social, humanitarian and environmental challenges.

The problem of picking out humpback whale songs underwater is particularly difficult to solve for several reasons. Underwater noise conditions can vary: for example, the presence of rain or boat noises can confuse a machine learning model. The distance between a recorder and the whales can cause the calls to be very faint. Finally, humpback whale calls are particularly difficult to classify because they are not stereotyped like blue or fin whale calls—instead, humpbacks produce complex songs and a variety of vocalizations that change over time.


A spectrogram (visual representation of the sound) of a humpback whale song in Hawaii.

We decided to leverage Google’s existing work on large-scale sound classification and train a humpback whale classifier on NOAA’s partially annotated underwater data set. We started by turning the underwater audio data into a visual representation of the sound called a spectrogram, and then showed our algorithm many example spectrograms that were labeled with the correct species name. The more examples we can show it, the better our algorithm gets at automatically identifying those sounds. For a deeper dive (ahem) into the techniques we used, check out our Google AI blog post.

Now that we can find and identify humpback whales in recordings, it allows us to understand where they are and where they are going—as shown by the animation below.


Since 2005, NOAA’s Pacific Islands Fisheries Science Center has deployed, recovered and collected recordings from hydrophones moored on the ocean bottom at 12 sites. On this map, you can see the spots where more whales were found by our classifier in orange and yellow.

In the future, we plan to use our classifier to help NOAA better understand humpback whales by identifying changes in breeding location or migration paths, changes in relative abundance (which can be related to human activity), changes in song over the years and differences in song between populations. This could also help directly protect whales by advising vessels to modify their routes when a lot of whales are present in a certain area. Such work is already being done for right whales, which are easier to monitor because of their relatively simple sounds.

The ocean is big and humpback whales are not the only ones to make noise, so we also started training our classifier on more species sounds (like the southern resident killer whale, which is critically endangered). We can’t see the species that live underwater, but we can hear a lot of them. With the help of machine learning, we hope that one day we can detect and classify a lot of these species sounds, giving biologists around the world the information needed to better understand and protect them.


A humpback whale breaching at the surface of the water. (Photo credit: Hawaiian Islands Humpback Whale National Marine Sanctuary.)

Reimagining the Google supply chain

Ever wonder how something like the Google Pixel starts out as a design and ends up in the palm of your hand?

To make products, like the Google Pixel, and all the technology that powers them—like Search, Gmail and YouTube—we rely on our supply chain. As we create more products and services for you, we also expand the reach of our supply chain to include new places, people and materials. Today, this includes more than 500 suppliers around the world who support our operations and manufacture hardware for devices and data centers.

As we work to continuously improve the way we design, source, produce, deliver, repair, and recover products, we wanted to share our commitment to create value along the entire supply chain. In our Responsible Supply Chain report, you’ll learn how we are working with individuals, communities and environments to do so. The report highlights our commitments and the progress we’ve made in areas like worker well-being, environmental impact and conflict minerals.

See the supply chain in action with immersive VR

If you want to learn more about where materials and products come from and the people and places that help create the devices we use in our daily lives, take a look at two new virtual reality project.  

In our Made by Me VR experience, you’ll see what it’s like to walk in a worker’s shoes for the day at a Flex supplier factory in Zhuhai, China. With a smartphone and Daydream View, Cardboard, or another VR headset, you’ll be immersed in a 360-degree environment of the factory where you can view the factory floor and scan the faces of workers during an impromptu break-time corner.  

Next up, in the Journey Of Gold, we will take you to the Democratic Republic of the Congo (DRC) to see how the Nyamurhale mine—which once sold illegally mined, taxed and smuggled gold—has now implemented systems to aid conflict-free gold sourcing.

Similar to these two stories, we envision a supply chain that equally values people's lives, the environment, and local communities. Through transparency and collaboration, our goal is to unlock new possibilities and advance the technology industry toward a more sustainable future. Learn more at sustainability.google/responsible-supply-chain/.

The Internet is 24×7. Carbon-free energy should be too.

Electricity is the fuel that allows our data centers to deliver billions of Google searches, YouTube views, and much more—every single day, around the clock. Our commitment to carbon-free energy should be around the clock too.

Today we published an inside look at the sources of Google's electricity around the globe, to gauge how we're tracking toward our long-term aspiration of sourcing carbon-free energy on a truly 24x7 basis. Our new discussion paper highlights how some of our data centers—like the one in Hamina, Finland—are already performing remarkably well on this front. The paper shares location-specific “Carbon Heat Maps” to visualize how well a data center is matched with carbon-free energy on an hour-by-hour basis. For Hamina, a heat map shows that 97 percent of the facility’s electricity use last year was matched with carbon-free sources.


Last year, 97 percent of our Finland data center’s electricity use was matched on an hourly basis with carbon-free sources.

The predominance of carbon-free energy at our Finland data center is partly due to Google’s purchases of wind energy in the Nordic region. Indeed, our large-scale procurement of wind and solar power worldwide is a cornerstone of our sustainability efforts, and has made Google the world’s largest corporate buyer of renewable energy. Last year we matched 100 percent of our annual electricity consumption with renewable energy purchases, and will continue to do so as we grow.

In many cases, we’ve partnered with local utilities and governments to increase the supply of renewable energy in the regions where we operate. For example, near our data center in Lenoir, NC, we worked with our local electricity supplier to establish one of the first utility solar purchase programs in the U.S. Solar alone, however, is unable to provide electricity around the clock. When the sun is shining, our Lenoir data center is quite carbon-free (indicated by the midday green ribbon in the Carbon Heat Map below), but at nighttime it’s more carbon-intensive; we plan to tackle this issue in the coming years by procuring additional types of carbon-free energy.


Last year, 67 percent of our North Carolina data center’s electricity use was matched on an hourly basis with carbon-free sources.

The Carbon Heat Maps demonstrate that there are times and places where our electricity profile is not yet fully carbon-free. They suggest that our 100 percent renewable energy purchasing goal—which relies on buying surplus renewable energy when it’s sunny and windy, to offset the lack of renewable energy supply in other situations—is an important first step toward achieving a fully carbon-free future. Ultimately, we aspire to source carbon-free energy for our operations in all places, at all times.

Creating a carbon-free future will be no easy feat, but the urgency of climate change demands bold solutions. Our discussion paper identifies several key actions that we and the rest of the world must take—including doubling down on renewable energy purchases in a greater number of regions—to achieve 24x7 carbon-free energy. We have our work cut out for us and couldn’t be more excited to push forward.

A breath of fresh air: Measuring air quality in Copenhagen

Healthy cities are important to everyone. And from a mother of an asthmatic child looking for the best way to get to the playground, to bike commuters and outdoor athletes finding the healthiest route for their trip, to city planners working to reduce unhealthy emissions, air quality information is crucial to making decisions in our daily lives. More detailed air quality insights are the goal of Project Air View, which kicked off today in Copenhagen, Denmark, in a partnership between the City of Copenhagen and Google.

Denmark’s National Center for Environment and Energy has estimated that around 550 Copenhageners die prematurely every year from pollution, and an even larger number suffer from related diseases; the yearly societal cost is estimated around 600 million euros.

Project Air View can help Copenhagen tackle this problem. It uses Google Street View vehicles equipped with scientific instruments that measure air quality at the street level. This creates a dataset which can map hyperlocal, block-by-block emissions and particle pollution. These measurements will be shared with scientists, the City Council, and ultimately, the public via interactive maps, all in an effort to tackle this well-known—and harmful—problem in big cities.

Today, Copenhagen has three stationary measuring points for air quality. We hope to complement the measurements from these fixed locations with our mobile Street View cars, enabling the City to measure air pollution in the City in significantly more detail. The air quality sensors measure nitric oxide, nitrogen dioxide, particulate matter, and ultrafine particulate matter.

We’ve enlisted the help of scientists from the University of Utrecht in The Netherlands to equip the car with the air quality equipment. They will also play an integral part in data validation and analysis. Aarhus University in Denmark will also contribute.

air view copenhagen

Equipping Street View cars with sensors to measure air quality

This is the next phase of our efforts to map air quality, after first mapping the City of Oakland and other California cities since 2015. We’ve also mapped high-resolution air quality data in London, and recently announced that we’re expanding to more places around the globe.

Project Air View is an example of how we can extend Google’s mission to air quality information, helping to reduce pollution and meaningfully impact people’s quality of life. And it’s one of several efforts aimed at applying technology to the world’s most pressing problems. We’re thrilled that Copenhagen is one of the first cities in Europe with a Street View car on the road measuring air quality. It speaks to Copenhagen’s forward-thinking approach to climate and environment, and we’re happy to contribute to that effort.

Air View is ready to expand to more places around the globe

Clean air is critical to life on Earth, yet over90 percent of the world’s population breathes polluted air. Over the past few years, we’ve been using a handful of Street View vehicles to gather air quality measurements, which can produce insights at the neighborhood level and can help cities become smarter and healthier.

Along with Aclima, we've been testing air quality equipment with the goal of fine-tuning their mobile-friendly air sensors to a point where they deliver accuracy comparable to laboratory-grade instruments.  After years of effort we’ve now achieved this goal. Today, we’re announcing that we will expand our air quality mapping to more Street View cars in more places around the globe. The locations are to be determined, but we have 50 air quality sensor-equipped Street View cars ready to hit the road.

During our initial research phase, Google and Aclima tested air quality equipment on a few Street View cars. Each car was installed with two sets of instruments: the first set contained laboratory-grade air quality reference instruments that are typically used for government air quality monitoring. This equipment is expensive and big, so it’s hard to deploy on a large number of vehicles. The second set had Aclima’s smaller, more mobile-friendly, air sensors that enable us to deploy in higher numbers. With both sets of equipment side-by-side, we've been able to validate their performance, and we’re now confident that the smaller Aclima sensors are ready to be deployed in 50 Street View cars.

Air View 1

Aclima’s sensor node in a Street View car.

This expansion builds upon work we’ve done in California over the past year. Our partners at the Environmental Defense Fund (EDF), University of Texas-Austin, and Aclima published a study showing that our mobile measurements can produce a map of air quality changing block by block. Scientists with Kaiser Permanente and EDF used the data to show a link between street-level air pollution and heart disease among the elderly. We began sharing the validated scientific air quality measurements with researchers, and will continue to make all of the street-level air quality data captured to date accessible with over 250 million measurements over four years and more to come with this expansion. Scientists and researchers are invited to request access to this data for air quality studies here.

Air View 2

Left: Black carbon particles come from burning fuel, especially diesel, wood and coal. Air quality data from Google and Aclima; analysis by Apte et al/EDF. Right: Air quality measurements in the San Francisco Bay Area region. Air quality data from Google and Aclima.

The measurements captured by these specially-equipped Street View cars around the world will show a snapshot of air quality at a moment in time, and can be used by scientists to combine with other data to develop air quality models. With this data, cities will be able to make more informed decisions and accelerate effort in their transition to a healthier city.

Source: Google LatLong

The more you know: Turning environmental insights into action

This week, thousands of leaders from cities, states, businesses, investors, and environmental organizations—including representatives from Google—will gather in San Francisco, CA at theGlobal Climate Action Summit to commit to raising the level of ambition in the fight against climate change. National governments around the world have committed to take action, but cities and businesses have an equally critical role to play in reaching a zero-carbon future. That's one reason we're excited about today's announcement of a new tool aimed at helping cities lower emissions.

Cities as diverse as New York, Berlin, Oslo, and Rio de Janeiro have committed to reducing their carbon footprint by 80 percent within the next 30 years. These cities rely on huge carbon data sets as a measuring stick to help identify where they may be able to cut emissions.  But many cities lack the resources to gather data such as building emissions, making it hard for them to set firm carbon commitments of their own.

The Environmental Insights Explorer (EIE), a new online tool we created in collaboration with the Global Covenant of Mayors for Climate & Energy (GCoM), is designed to make it easier for cities to access, and act upon, new climate-relevant datasets.  By analyzing Google’s comprehensive global mapping data together with standard greenhouse gas (GHG) emission factors,  EIE estimates city-scale building and transportation carbon emissions data, as well as renewable energy potential, leading to more globally-consistent baselines from which to build policies, guide solutions, and measure progress.

Introducing The Environmental Insights Explorer

To date, more than 9,000 cities have made commitments to comply with the Paris Agreement, which presents a formal plan and timeline to phase out reliance on fossil fuels. But less than 20% have been able to complete, submit or monitor greenhouse gas inventories.

The process for building out a baseline emissions inventory can take hundreds of thousands of dollars and months or even years. “The vast majority of cities aren’t in the position to finance a process that will take time and might be cost prohibitive, especially the small to medium cities in developing areas of the world. And that’s where most of the action will take place in relation to the Paris Agreement on climate change,” explains Amanda Eichel, the executive director for the global secretariat at GCoM, an international alliance of nearly 10,000 cities and local governments committed to fighting climate change.

With EIE, data sets that once required onsite measurements and many months to compile can now be assessed virtually, reducing cost and time investment that prevents cities from taking action.

Data packaged to prompt action

On the EIE site, data is freely available in four categories: building emissions, transportation emissions, energy offset potential, and 20-year climate projections. Clicking on “Building emissions,” for example, brings up detailed maps visualizing the emissions impact for both homes and non-residential buildings.

In each category, you can drill down into more specific statistics, including percentage breakdowns of emissions, the time period from which the data was culled, key assumptions made. You can also find links to other critical information, such as ways to reduce emissions.


Emissions data gets more specific the deeper you get into the site.“This tool will provide us with much more precise data on the flow of transport emissions and the potential of the City to generate solar energy,” says Buenos Aires Mayor Horacio Rodríguez Larreta. “It is a key input to design policies that reduce emissions and make Buenos Aires a smarter, greener and more sustainable and resilient city.”

In addition to helping policy makers, planners, and researchers set city-wide emissions policies, the data can inform specific projects like new investments in solar, public transit, or mobility alternatives to reduce vehicle traffic. For example, a city could track a new transportation line’s potential impact on the city’s emissions profile before deciding whether to move forward and scale the project. Or a city could explore how transitioning some percentage of short car trips to bicycle trips would lower the overall carbon footprint. "Now we can bring data analytics to conversations about renewable energy and show people that they’re able to generate enough solar power for their entire city," says Brad Petry, Head of Data Analytics, Victorian Centre for Data Insights. Victoria's state government has set targets for 25 percent renewable energy by 2020 and 40 percent renewable energy by 2025.


EIE estimates total solar potential for rooftops to show how much renewable power could be generated, helping cities evaluate ways to reduce overall building energy emissions.

Filling an information gap, collaboratively

EIE arose from a decade’s worth of climate-related projects at Google, including Project Sunroof, a tool that measures rooftop solar energy potential, and Earth Engine, a platform for geospatial analysis. Collectively these projects in conjunction with other Google data sources like building and transportation data, were pulled together to reveal valuable insights about cities' carbon impact—information that we realized could play a critical role in encouraging action by policy makers, city officials, and others.  But to be effective, the information needed to be packaged for easier absorption and more importantly, action.

We started by partnering with GCoM, founded by global city networks and supported by Bloomberg Philanthropies and the European Union,  which has been gathering the same comprehensive data we wanted to surface in our tool through other sources and methodologies. GCoM also has detailed knowledge of the intricacies of environmental policy and the political hurdles hindering change and action.

We shared our data inventories with one another, and GCoM helped connect us to different cities around the world to get their input on what they’d find most useful to help reach their emission goals. The methodology used to source, aggregate and distill the EIE data sets can be reviewed on the site. To ensure quality, we initiated a rigorous quality-assurance process months before launch.


A work in progress

We're introducing EIE in beta today, covering a handful of pilot cities including Melbourne, Australia; Buenos Aires, Argentina; Victoria, Canada; Pittsburgh, Pennsylvania; and Mountain View, California. But in time, we plan to make this environmental information available to thousands of cities, towns, and regions around the world. As more cities use this data, and as science evolves, we plan to iterate and expand on the tool, methodologies and datasets.  

Even providing thousands of cities with comprehensive, action-oriented datasets is just one piece of the emissions mitigation puzzle. Still, we’re excited to take this first step today with GCoM and pilot cities on a journey to accelerate global ambition and action toward a low-carbon future.

Learn more about Google’s other sustainability efforts in our 2018 Environmental Report and on sustainability.google.

Source: Google LatLong