News Article | April 22, 2015
Forests are often described as the lungs of our planet. They’re home to countless animal and plant species, provide humans with oxygen and food, and help stave off the effects of global warming. Yet we as a species are constantly imperiling these vital ecosystems. An estimated 46 to 58 thousand square miles of forest are cut each year. At the current rate of deforestation, the world’s rain forests could be obliterated within just 100 years, according to National Geographic. Now environmentalists are turning to artificial intelligence and satellites to combat forest destruction in real time—that is, in time to stop it. Last year the World Resources Institute launched a website called the Global Forest Watch to provide up-to-date information about the state of forests worldwide. The trouble is that the site has often been stuck reporting deforestation after the fact. Now one startup is working to give the organization the tech to spot deforestation risks before they cause irreversible harm. Orbital Insight, founded by former Google and NASA robotics and artificial intelligence expert James Crawford, plans to collect satellite imagery of tropical forests to track changes over time. The company will set its machine vision algorithms loose on the images to detect small alterations in the landscape that humans might miss but that could indicate illegal logging. At first, the visual impact of building a new logging road in a huge forest is so small that the human eye will have trouble noticing any difference, Crawford explains. But as loggers use that road to clear more and more trees, the effects become more and more obvious. If that road can be spotted before much logging has actually been done, activists and governments stand a much better chance of preventing more damage. Crawford says Orbital Insight will draw on a wide range of satellite image sources and analyze the data using an artificial intelligence technique known as deep learning — the same approach that companies like Google and Facebook are using for image recognition and other tasks. He says the company may also pull in other data, such as temperature and rainfall, as part of the analysis. The project is still in its earliest stages, but Orbital Insight has done this sort of work for private customers before. For example, it has helped hedge funds determine which big box stores have the most customers by comparing the number of cars that routinely appear in each company’s parking lots. It’s estimated the world’s oil reserves by analyzing changes in the shadows cast by the floating lids of oil tanks around the world. Crawford says the Orbital Insight team wanted to find ways to apply its technology to other problems that might benefit humanity, and they were all impressed with the work that WRI has been doing. Ultimately, Orbital Insight and environmentalists don’t have the ability to stop logging. That responsibility will fall on governments and citizens. But the job might get a little easier if the people working to save forests can anticipate the damage before it’s done, rather than being left to do damage control after it’s too late.
News Article | April 3, 2015
New high-resolution maps of international woodland reveal that the world lost 18 million hectares of forest—the same area as that covered by Oklahoma—to wildfires, deforestation, and development in 2013 The maps, such as the one shown above, were created by Global Forest Watch in a collaboration between University of Maryland an Google, bringing together more than 400,000 satellite images from by NASA’s Landsat mission. An incredible 34 percent of all that loss was found in Canada and Russia, losing 2.5 million hectares and 4.3 million hectares respectively. Their northern boreal forests suffered at the hands of large-scale wildfires. As Motherboard points out, such losses could be part of a viscous global cycle. There’s evidence to suggest that wildfires become more frequent as global temperatures rise; as trees burn, they release large quantities of carbon into the atmosphere that would otherwise remain locked inside them, exacerbating the effects of climate change. Sadly, loss may lead to more accelerated loss. [Global Forest Watch via Motherboard]
News Article | July 1, 2014
You probably ate some palm oil today. Or drank some. Or rubbed some on your skin. If you cleaned the house, then you probably scrubbed your floors, washed your windows, and then freshened the air with products containing palm oil. The stuff is in almost half of all products on U.S. supermarket shelves, so you’d have to be a pro at ingredient label hopscotch avoid it. But you might want to try. Demand for palm oil (which comes from, you guessed it, the oil palm) has caused mass deforestation in Malaysian and Indonesian rainforests. These are some of the most biologically diverse areas in the world, and home to endangered Sumatran tigers, orangutans, and thousands of lesser-known species. On top of the environmental concerns, some farmers use fire to clear these forests, which has caused civil strife and public health concerns in cities like Singapore, which have been choked by haze and smoke. Recently, a palm oil industry group called Roundtable for Sustainable Palm Oil (RSPO for short, see sidebar below), teamed up with conservationists to create a map intended to make the palm oil supply chain more transparent. The map shows land concessions granted by governments, deforestation (and reforestation), active fires, and plantations that have been certified as sustainable. Some parts of the map are only accessible to RSPO member companies, which can use it to get their products certified as sustainable. But the public also has access to some of the data, which means consumers, conservationists, and government officials can now monitor the industry more actively. “This goes towards something we have been pushing for for a long time, which is traceability,” said Joao Talocchi, a palm oil campaigner for Greenpeace. “For the first time, this is letting the public and companies monitor the RSPO policies.” Even environmental stalwarts like Greenpeace believe the crop can be reformed. “We’re not against palm oil, but against palm oil not created in a sustainable way,” Talocchi said. To make the new map, RSPO gave their data to the non-profit World Resources Institute (WRI), which added it to their mapping platform, Global Forest Watch. Built with geographic analysis software, Global Forest Watch lets you view areas where palm oil is being harvested (both sustainably and otherwise) on top of other data, such as active forest fires, land cover type, cumulative forest loss since 2001, and more. RSPO has mandated that palm oil only be grown on secondary forest cleared without using fires, but there’s evidence that doesn’t always happen. “The main issue that we’re looking at is the link between palm oil concessions and deforestation,” said Elizabeth Baer, a spokesperson for WRI. So, let’s say you’re an investigator for an environmental NGO looking to see if there are any fires or deforestation on any RSPO certified concessions. Go to the map’s landing page and click the leftmost circle (the “Beta users” referred to in the other two circles are select RSPO members). The map loads over Southeast Asia, with the active layer showing cumulative deforestation since 2002. Mess around with the sliders and radio buttons a bit if you want to get acquainted, then click on the far-left menu titled “Select Contextual Maps.” This will open a sub-menu that lets you add land cover layers. Stay away from this for now (the extra colors can camouflage the layers you’re really interested in) and click on another sub-menu below, titled “Land Use”. In response to a growing outcry that palm oil was wrecking the rainforest, in 2004 nearly 50 palm oil companies and organizations signed the charter for the Roundtable for Sustainable Palm Oil . The group includes representatives from every link in the palm oil supply chain: growers, processors, manufacturers, retailers, banks, as well as environmental and human-focused NGOs. Its stated mission is to make sustainability the norm throughout the the entire palm oil supply chain.The RSPO developed a number of criteria for sustainable growth, which include things like not cutting down old growth forest, not burning peat, not treating employees like animals, and being transparent about all these things so the RSPO can make sure they’re following the rules. Members can apply for RSPO certification and have their business audited to make sure they’re running a clean shop. But, the process doesn’t always work. Only RSPO members can get certified, but simply being a member does not ensure good behavior. In 2009, consumer goods giant Unilever cut all ties with RSPO member PT Smart—a subsidiary of the largest palm oil producer in Indonesia—after confirming reports from Greenpeace that PT Smart was burning down forests and draining peat bogs. This and other incidents have led companies like Unilever and Nestle to work outside the RSPO to deliver sustainable products. Clicking the the second box will show you all the RSPO certified palm oil concessions. Now you can pan and zoom to see where palm oil has caused deforestation. Try clicking on the land cover layer (turning up the transparency helps), and you can whether that deforested land was primary—old growth—or otherwise. Or go back to the original main menu (titled “Select Forest Change Maps”) and you can see if any of those concession areas are on fire. Navigating between menus can be a pain, there are some missing layers (like the locations of palm oil processing plants), and the color scheme can be overwhelming. But, WRI is still making improvements. Among the most exciting will be a layer that shows the ground-truthed location of every palm oil plantation. Fred Stolle is a geographer for WRI, and one of the architects of Global Forest Watch. He says farmers don’t always break ground where they’ve been granted their concessions, so WRI hired a group of analysts to locate plantations by poring over satellite imagery. (Even though oil palm plantations have a distinct pattern, automatic recognition programs have a hard time telling them from the surrounding forest in satellite imagery, says Stolle.) Only about 12 percent of the world’s palm oil is grown according to RSPO’s certification standards. But, because oil palms are productive and easy to grow, they have the potential to be the poster crop for sustainability in the globalized economy. “Palm oil is not a bad versus good story,” said Stolle. For example, he says, the crop has helped drive economic growth in Malaysia’s economy. “It can be great for the tropics. If the companies would plan it right, it could be a complete success story,” said Stolle. The map could play an important role by allowing the final link in the supply chain—consumers—to trigger accountability that reaches all the way back to the source. “You can see that what you eat and consume… is linked to an environmental impact a 1,000 miles away,” said Talocchi. A tool like this, he says, allows anyone to make choices that affect the way that business is done.
News Article | April 2, 2016
At the turn of the 20th century, roughly 100,000 wild tigers roamed the forests of Eurasia. Today, that number has dwindled to a global population of under 3,500 individuals, driven to the brink of extinction by anthropogenic pressures like poaching and the loss of 90 percent of their historic range to human encroachment. But the good news is that this iconic predator is finally beginning to rebound in some areas, in the wake of concerted efforts to restore its habitat. The extent of this slow bounceback is analyzed in detail in a study published on Friday in Science Advances, which collates satellite observations of forest loss from 2000 to 2014 in 76 regions earmarked for tiger conservation across 13 countries including China, India, Indonesia, and Russia. The team, led by University of Minnesota conservation scientist Anup R. Joshi, demonstrate that this space-down Big Data approach is proving to be instrumental in tracking the recovery of tigers with precision detail and in near-real time. Satellites allow scientists to see exactly which stretches of the animal’s limited range are imperiled the most, and they can use that global view to make estimates about the health of tiger populations and to direct resources to threatened areas. “Our analysis wouldn't have been possible without free access to satellite data and Google Earth Engine [Google cloud] for computation,” Joshi told me via email. “[M]ore governments and organizations are providing free data to the public,” he continued. “With advances in technology, more data on Earth observation is coming in every day than we can analyze. Machine learning algorithms are becoming more reliable, which will change the way we monitor forest and other natural resources in coming years, to quickly analyze large data sets and identify trouble spots.” In particular, this study focused on the progress toward the “Tx2” goal, set by the governments of tiger-inhabited countries in 2010, which aims to double the worldwide tiger population to 6,000-plus individuals by 2022. In a pleasant twist, satellite imagery sourced from Google Earth Engine and Global Forest Watch revealed that forest loss was not as bad as anticipated, with most of the trouble spots confined to only ten of the 76 prioritized landscapes, according to the new study. This means that reaching the Tx2 goal is still on the table, so long as conservationists continue to ensure tiger habitats are restored and protected. That said, the data confirmed that the conversion of tiger habitat into farming and urban communities continues to plague the cats—particularly with regards to oil palm plantations. “Expanding oil palm plantations in the tiger range countries has been one of the main causes of forest conversion and habitat fragmentation and loss,” Joshi said. “In Indonesia alone, more than 4,000 square kilometers or 1,544 square miles of forests have been allocated for oil palm concessions.” “Forest clearing has a direct effect on tiger habitat and its prey species. Tigers need large areas to meet their food, shelter, and mate needs. Others threats are growing infrastructures, such as roads and cities to support the increasing human population.” Indeed, tiger territory frequently overlaps with dense rural human populations in nations with some of the fastest-growing economies on Earth, such as China and India. That is bound to be disruptive to their habitat and breeding. But as our space-down view of the world becomes sharper, it will be easier to identify which areas require the most urgent attention. As an example, Joshi’s team zoomed in on satellite footage of the Khata and Basanta corridors in Nepal, which are forested links between larger wildlife reserves. The results revealed a stark contrast: the Khata corridor is thriving due to forestry stewardship efforts and anti-poaching campaigns, while the Basanta corridor has been crippled by increasing human encroachment. It is this kind of precision analysis of specific regions that will help tiger advocates to root out where conservation efforts are failing, and further intervention is needed. The best part is that this technology is not accessible to scientists alone; anyone can go to Google Earth Engine and Global Forest Watch to keep an eye on critical tiger habitats. “Local communities will be able to see how their government officials are managing forest resources, protected areas,” Joshi said. “Likewise, donors and advocacy groups would be able to monitor projects thus more transparency and accountability. When people have interactive tools which are simple to use, more people will participate in discussing issues and coming up with solutions.”
In the aftermath, automobile, electronic, chemical, and retail sectors worldwide, which relied on Japanese manufacturers for parts and materials, suffered massive shortages. Few affected companies knew enough about the complex Japanese supply chain to respond to such an immense disruption. "Companies hadn't been keeping good mapping records of where their suppliers were, or where their suppliers' suppliers were, so they asked us to deploy [Sourcemap] inside their supply chains," Bonanni says. "All of a sudden all these maps—the first to show products all the way from raw materials to consumers, every step of the way—became a critical tool for manufacturing companies." Motivated by this incident, Bonanni launched Sourcemap commercially so companies could keep better tabs on their supply chains. Today, dozens of pharmaceutical firms, food producers, and clothing and electronics companies use Sourcemap. The software, which operates in the cloud, gives companies a visual map of all connected global supplier locations for their products. The software also pulls news feeds on significant political events or environmental disasters to alert companies to potential interruptions. Sourcemap has also continued its original mission of increasing the transparency of supply chains for consumers. The company offers an open-source platform for anyone to publish supply chain maps. And today Sourcemap announced what Bonanni calls "the most ambitious supply-chain transparency project ever." Partnering with Imaflora, a certifier in Brazil, Sourcemap is deploying an online tool, called Origens Brasil, that allows consumers in the country to enter serial numbers of numerous products and see exactly where the products came from. Moreover, Bonanni says, consumers can see if certain products—such as cocoa, palm oil, and paper—come from producers that are contributing to deforestation of the Amazon rainforest. To do so, the software looks at Global Forest Watch satellite imagery for any noticeable decrease in leaf cover where all ingredients of a product are made. The tool could also be attractive for multinational corporations pledging to support only producers that do not contribute to deforestation, Bonanni says. "This is a totally new level of confidence that you can have when you buy things," he says. Supply chains are surprisingly complex systems of organizations, people, and resources that transform raw materials into finished consumer products. Major supply chains, such as those for mobile phones, can sometimes include hundreds or thousands of direct and indirect suppliers of hardware, components, parts, and software. In fact, Bonanni says, when most companies see their entire supply chains mapped instantly for the first time, they're quite shocked. "They are struck by the amount of zigzagging that is going on," he says. One of the most complex supply chains Bonanni has seen was the first one he tried mapping himself 10 years ago, as part of a research project through the MIT Media Lab and MIT Center for Civic Media: that of the computer chip. One chip, he says, consists of about 50 different materials—including some rare-earth elements—produced from at least that many countries. He never finished the map. "The computer chip will travel dozens of times around the Earth, if you add up all the paths of all of its subcomponents, before it's finally created," he says. "What's amazing is you have millions of people to make a computer chip, and billions of people now own computer chips. So if you draw the whole supply chain of semiconductors, it's a social network that includes almost everyone on the planet." Traditionally, companies hire consultants to map their supply chains. This is a labor-intensive process, where consultants track down parts and materials, draw out maps with computer tools, and analyze for risks and opportunities. Sourcemap, however, is automated, fast, and visually appealing, Bonanni says. Using Sourcemap, companies upload product and supplier data, and revenue from each product line. Surveys are sent to suppliers to provide missing information, such as where their raw materials and subcontractors are located. Supply chains appear as color-coded dots (supplier locations) and lines (shipment paths) connecting across a map. For each supplier, Sourcemap reveals the inventory count and calculates the money the company would lose if that location went offline for any reason, accounting for time to find a new supplier and how long inventory will last. Companies can also modify supply chains to plan for new products, customers, or suppliers. The software also employs predictive analytics to analyze real-time news feeds about disasters, corruption, local conflicts, or climate change, and alert companies to find alternate routes to avert disaster. It will track customer demographics in different locations to help companies decide on branching out to new markets. "[Companies] see whether they need to think strategically about shifting the company direction in a big way to avoid running into a wall, whether that means climate change is making the crop that you rely on harder to grow or the demographics of your customer base is shifting," Bonanni says. "It's not until you see the entire operations mapped on one screen that you can start to make those decisions." According to Bonanni, companies are beginning to share their data to make supply chains—for products such as minerals, apparel, or cocoa—more transparent within industry. Under increased public scrutiny, he adds, companies are becoming more socially conscious of their supply chains, as well. In 2010, when Sourcemap was still a research project, Stonyfield Farm mapped their entire ingredients supply chain for the public. In 2013, Sourcemap partnered with Mars Chocolate on Vision for Change (V4C), an initiative to improve cocoa productivity and open up direct connections with cocoa farmers to improve their livelihoods. Last year, Fairphone, a company that aims to make environmentally sustainable smartphones, became the first electronics company to publish a Sourcemap of its supply chain, which included more than 300 parts from companies from Japan to North America. Origens Brasil is another step toward Sourcemap's mission of greater supply chain transparency, Bonanni says. "The goal for us is to enable that type of visibility, enable companies to see beyond their walls," he says. "For financial gain, but also for social good, to make sure they're not using up natural or social resources faster than they can be replenished." Explore further: What smart companies can learn from Apple's supply chain woes