Russell Wilson’s startup Tally powers new real-time predictions game for L.A. Rams, Seahawks’ big rival

Russell Wilson’s startup Tally powers new real-time predictions game for L.A. Rams, Seahawks’ big rival

5:08am, 3rd September, 2019
Seattle Seahawks quarterback Russell Wilson, right, with Jason LeeKeenan, CEO of Tally, in Seattle in 2018. (GeekWire Photo / Kevin Lisota) When the Seattle Seahawks take on the Los Angeles Rams on Oct. 3 this season, fans might want to predict how many touchdowns quarterback Russell Wilson will throw against his division rival on that day. The best way to do so could be through a new mobile experience from the NFL team that is powered by , the startup that was founded by Wilson. The Rams launched “Pick’em” for use during a pre-season game against the Oakland Raiders. The intention is to engage fans to make real-time predictions as the action unfolds on the field. Fans, playing on the web or through the Rams’ mobile app, earn points for every correct prediction and those over 18 can compete for prizes such as game tickets, field passes and autographed merchandise. Tally is a free-to-play predictions platform, not a gambling app. But the move by the Rams, along with a , signals what’s ahead with the eventual spread of legalized sports betting in the wake of a 2018 that overturned the Professional and Amateur Sports Protection Act. It’s all poised to change how we watch and interact with live events. Seattle-based Tally, which employees 14 people now, is an , the company Wilson helped launch in 2017 as a celebrity content app. TraceMe shut down in 2018 and the business pivoted to the sports prediction model. Wilson was touting Tally in February. “We believe that real-time predictive gaming experiences are going to be the critical components of engaging in live sports in the years to come,” Tally CEO Jason LeeKeenan told GeekWire. “We are positioning Tally to be the leading technology provider behind this evolution.” The Tally app, showing, from left, phone authentication, dynamic odds, and a real-time leaderboard. (Tally screen shots) According to its website, Tally white labels its user interface, custom branding it for any property looking to create such content. The Rams are the first NFL team to partner with Tally. LeeKeenan said other partnerships are in the works, but he wasn’t ready to announce whether the Seahawks might be one of those teams. reported that Tally worked with the NBA’s Portland Trail Blazers and the NHL’s St. Louis Blues on similar games earlier this year. The Rams plan to use the mobile experience to present a mix of game-specific questions and micro-outcomes, according to the team’s news release. “Which team completes a passing play of 30+ yards in the opening half?” or “Which of these players racks up 10+ rushing yards first?” are example of questions posed to fans. Those playing can can track their success throughout a game and rankings are updated in real time as results are tallied live. Point values increase as the game progresses. “We are thrilled to bring our fans closer to the action with an engaging second-screen experience,” Marissa Daly, Rams VP of media, said in a statement. “We feel that our free-to-play predictions game will be a fun way for fans to compete against one another while watching their Rams compete on the field.” The Rams have leapfrogged the San Francisco 49ers to emerge as the Seahawks’ most heated division rival over the past couple seasons. Surely Seattle’s star QB will be more engaged with winning games on the field than worrying about predictions being generated in an app built by his company. Regardless, LeeKeenan makes it sound like Wilson has already won. “What can I say? Russell is a great entrepreneur and we hope all sports teams will be using our technology one day,” LeeKeenan said.
Washington Hyperloop slims down its pod racer for Elon Musk’s next big contest

Washington Hyperloop slims down its pod racer for Elon Musk’s next big contest

12:51am, 13th May, 2019
Washington Hyperloop team members show off their Husky spirit at an on-campus unveiling of this year’s pod racer. Veteran team member Mitchell Frimodt peeks out from within the pod’s carbon composite shell, while the guts of the racer are on display on a table at left. (Margo Cavis Photo) Could this year be the year for ? For the fourth time, the students on University of Washington’s pod-racing team are taking aim at the top prize in tech titan Elon Musk’s competition, and this time they’ve got their racer down to fighting weight. This year’s purple pod racer, which looks like a cross between a bobsled and a miniaturized bullet train, was unveiled Friday night at UW’s Husky Union Building. “Our pod this year is about 60 percent of the weight of last year’s pod, with the same propulsion specs,” engineering senior Mitchell Frimodt, one of the veterans on the Hyperloop team, told GeekWire. “That’s our performance boost.” Propulsive oomph per pound is a key factor in what’s become an annual tradition that plays out at SpaceX’s headquarters in Hawthorne, Calif. This year, Washington Hyperloop and a dozen other collegiate teams are . Competitors will show off the racers they’ve built, and the best of the pack will face off in time trials conducted in a mile-long tube that’s been built just across the street from SpaceX’s rocket factory. The fastest team wins. And in the previous three competitions, the fastest team has been WARR Hyperloop from the Technical University of Munich in Germany. This year, Munich’s student engineers are racing under a different team name — — but they’re expected to be every bit as formidable. “We’re looking to give them a run for the money,” Frimodt said. Frimodt is speaking figuratively: There’s no prize money as such, but the Hyperloop contest gives those who do well an enviable spotlight in the engineering world. Frimodt said some of Washington Hyperloop’s alumni are now working at SpaceX, Tesla, Apple and other cutting-edge ventures — including Musk’s very own tunneling venture, . “One of our old business guys is actually with the Boring Company currently, down in L.A.,” he said. “I haven’t heard exactly what he’s up to down there, but that’s pretty cool. We’ve had one alum go on to .” Musk as a way to short-circuit traffic snarls within and between urban areas. The original concept called for pods to travel through low-pressure tubes at near-supersonic speeds, cutting the travel time between the San Francisco Bay Area and Los Angeles to about a half-hour. To cite another example, a Pacific Northwest Hyperloop could . Since then, the intercity Hyperloop concept has given rise to commercial ventures that don’t involve Musk, including and . For now, Musk is concentrating on the Boring Company’s somewhat less speedy tunnel travel concept, which is known as the Loop. Loop projects are in various stages of development in , , and the . Frimodt said most of the 40 or so members of the Washington Hyperloop team see their involvement as a way to exercise their general engineering skills, rather than creating a commercially viable Hyperloop pod. “For us, it’s more on pushing the edge of technology. … Our focus is very much on the competition,” he said. Only about half of the team’s members are veterans of past Hyperloop contests. Thanks to the team’s previous successes, including an , Washington Hyperloop’s organizers received “a ton of applicants” seeking to join the team last fall, Frimodt said. So what’s the secret sauce for this year? It’s not necessarily the propulsion technology: The team is going with a cold-gas thruster system that’s similar to the one that won them fourth place in last year’s competition. Basically, pressurized nitrogen blasts through a rocket-like nozzle to shoot the pod down the Hyperloop track. “It’s a rocket without the combustion,” Frimodt explained. One big change from last year has to do with weight reduction. This year’s team members were rigorous about enforcing weight budgets for each of the pod’s components, and were lots more liberal about using lightweight carbon composites rather than metal parts. Another change has to do with advance preparation, particularly when it comes to the software side of the project. Last year, “people were working on our code on the day of the competition,” said Fedor Paretsky, a UW junior in applied physics who works on the control and power subsystem team. This year, the Washington Hyperloop team is ahead of the game, development-wise. And one of the reasons for that has to do with one of the team’s sponsors, a California-based tech company named , which is providing the team with financial support as well as help with an open-source programming language for cloud-based applications known as . “Ballerina acts as our data relay,” Paretsky said. “It relays data every 25 milliseconds. … If we decide that something’s going wrong, or we need to do an emergency stop, Ballerina ensures that the pod is communicated with, efficiently.” Ballerina also serves to coordinate software development. “Anybody can literally just pull the image [for the pod’s software] off the cloud,” he said. In addition to WSO2, Ballerina and various UW departments, the team has recruited ranging from Boeing to Pagliacci Pizza (and it doesn’t take an engineer to guess what Pagliacci is contributing). Their logos are all emblazoned on what Washington Hyperloop hopes will be this year’s top pod. It could be a nail-biter: Last year’s competition at SpaceX saw the WARR team set a new world record for Hyperloop travel, with a . Can Washington Hyperloop improve on that mark? “We’re not 100 percent sure yet, until we get a final pod weight based off everything we put on there, and do our static-fire testing to characterize our propulsion system,” Frimodt said. “But it’s probably looking like somewhere from 260 miles an hour to 300 miles an hour.”
Washington Hyperloop slims down its racing pod for Elon Musk’s next big contest

Washington Hyperloop slims down its racing pod for Elon Musk’s next big contest

1:49pm, 11th May, 2019
Washington Hyperloop team members show off their Husky spirit at an on-campus unveiling of this year’s pod racer. Veteran team member Mitchell Frimodt peeks out from within the pod’s carbon composite shell, while the guts of the racer are on display on a table at left. (Margo Cavis Photo) Could this year be the year for ? For the fourth time, the students on University of Washington’s pod-racing team are taking aim at the top prize in tech titan Elon Musk’s competition, and this time they’ve got their racer down to fighting weight. This year’s purple pod racer, which looks like a cross between a bobsled and a miniaturized bullet train, was unveiled Friday night at UW’s Husky Union Building. “Our pod this year is about 60 percent of the weight of last year’s pod, with the same propulsion specs,” engineering senior Mitchell Frimodt, one of the veterans on the Hyperloop team, told GeekWire. “That’s our performance boost.” Propulsive oomph per pound is a key factor in what’s become an annual tradition that plays out at SpaceX’s headquarters in Hawthorne, Calif. This year, Washington Hyperloop and a dozen other collegiate teams are . Competitors will show off the racers they’ve built, and the best of the pack will face off in time trials conducted in a mile-long tube that’s been built just across the street from SpaceX’s rocket factory. The fastest team wins. And in the previous three competitions, the fastest team has been WARR Hyperloop from the Technical University of Munich in Germany. This year, Munich’s student engineers are racing under a different team name — — but they’re expected to be every bit as formidable. “We’re looking to give them a run for the money,” Frimodt said. Frimodt is speaking figuratively: There’s no prize money as such, but the Hyperloop contest gives those who do well an enviable spotlight in the engineering world. Frimodt said some of Washington Hyperloop’s alumni are now working at SpaceX, Tesla, Apple and other cutting-edge ventures — including Musk’s very own tunneling venture, . “One of our old business guys is actually with the Boring Company currently, down in L.A.,” he said. “I haven’t heard exactly what he’s up to down there, but that’s pretty cool. We’ve had one alum go on to .” Musk as a way to short-circuit traffic snarls within and between urban areas. The original concept called for pods to travel at near-supersonic speeds, cutting the travel time between the San Francisco Bay Area and Los Angeles to about a half-hour. To cite another example, a Pacific Northwest Hyperloop could . Since then, the intercity Hyperloop concept has given rise to commercial ventures that don’t involve Musk, including and . For now, Musk is concentrating on the Boring Company’s tunnel travel concept, which is known as the Loop. Loop projects are in various stages of development in , , and the . Frimodt said most of the 40 or so members of the Washington Hyperloop team see their involvement as a way to exercise their general engineering skills, rather than creating a commercially viable Hyperloop pod. “For us, it’s more on pushing the edge of technology. … Our focus is very much on the competition,” he said. Only about half of the team’s members are veterans of past competitions. Thanks to the team’s past successes, Washington Hyperloop’s organizers received “a ton of applicants” seeking to join the team last fall, Frimodt said. So what’s the secret sauce for this year? It’s not necessarily the propulsion technology: The team is going with a cold-gas thruster system that’s similar to the one that won them fourth place in last year’s competition. Basically, pressurized nitrogen blasts through a rocket-like nozzle to shoot the pod down the Hyperloop track. “It’s a rocket without the combustion,” Frimodt explained. One big change from last year has to do with weight reduction. This year’s team members were rigorous about enforcing weight budgets for each of the pod’s components, and were lots more liberal about using lightweight carbon composites rather than metal parts. Another change has to do with advance preparation, particularly when it comes to the software side of the project. Last year, “people were working on our code on the day of the competition,” said Fedor Paretsky, a UW junior in applied physics who works on the control and power subsystem team. This year, the Washington Hyperloop team is ahead of the game, development-wise. And one of the reasons for that has to do with one of the team’s sponsors, a California-based tech company named , which is providing the team with financial support as well as help with an open-source programming language for cloud-based applications known as . “Ballerina acts as our data relay,” Paretsky said. “It relays data every 25 milliseconds. … If we decide that something’s going wrong, or we need to do an emergency stop, Ballerina ensures that the pod is communicated with, efficiently.” Ballerina also serves to coordinate software development. “Anybody can literally just pull the image [for the pod’s software] off the cloud,” he said. In addition to WSO2, Ballerina and various UW departments, the team has recruited ranging from Boeing to Pagliacci Pizza (and it doesn’t take an engineer to guess what Pagliacci is contributing). Their logos are all emblazoned on what Washington Hyperloop hopes will be this year’s top pod. It could be a nail-biter: Last year’s competition at SpaceX saw the WARR team set a new world record for Hyperloop travel, with a . Can Washington Hyperloop improve on that mark? “We’re not 100 percent sure yet, until we get a final pod weight based off everything we put on there, and do our static-fire testing to characterize our propulsion system,” Frimodt said. “But it’s probably looking like somewhere from 260 miles an hour to 300 miles an hour.”
Fast-growing sales tech startup Highspot leases big Seattle waterfront office with room for 800 people

Fast-growing sales tech startup Highspot leases big Seattle waterfront office with room for 800 people

5:59pm, 7th May, 2019
The World Trade Center East building. (Highspot Photo) will soon have a new spot to call home. The startup that builds artificial intelligence-powered sales software has leased two floors with options to take more at the World Trade Center East building near the Seattle waterfront. The 55,000-square-foot space, which will be ready around the end of the year, will have room for roughly 450 people. Highspot CEO Robert Wahbe. (Highspot Photo) Today, Highspot has about 200 employees and expects to hit 300 when it moves into the space. The company will hold on to its existing offices in Seattle, giving it a total footprint of more than 90,000 square feet and capacity for roughly 800 employees. Robert Wahbe, co-founder and CEO of Highspot, said the company is experiencing explosive growth in revenue and other key business metrics, and it is hiring fast to keep up. “We are growing more than 100 percent per year across all the normal business metrics and growing more than 100 percent in our headcount,” Wahbe said. “Given how competitive the environment is we are very focused on attracting and developing world-class people.” Last year, Highspot landed a to power its rapid growth. The company has raised more than $64 million in its lifetime. Wahbe called his company the fastest-growing tech startup with fewer than 1,000 employees in the area. He came to that conclusion by looking at headcount growth numbers on LinkedIn of companies in the index of the top Pacific Northwest startups. Highspot’s customer base is growing 300 percent year-over-year, Wahbe told GeekWire last year, adding to a big-name stable of customers that includes Amazon, Dropbox, Uber, Lyft Twitter, Zillow, Airbnb and SAP. A finalist for at the 2019 GeekWire Awards, Highspot equips sales teams with artificial intelligence-infused technology to improve how they have conversations with prospective buyers. Its “sales enablement platform” is a sales playbook of sorts, analyzing hoards of internally-produced information — historical data; marketing presentations; case studies; data sheets, etc. — and then applying AI to optimize the selling process. Highspot also provides communication and analytics tools with a goal of helping marketing and sales teams better collaborate. Highspot’s future office space. (Highspot Photo) The concept of bringing sales and marketing teams together has been around since the beginning of the modern office, but the technology hasn’t been there. That all changed around 2010, as mobile technology, AI and software-as-a-service innovations progressed rapidly. Since then, the category has taken on a renewed importance, Wahbe said. “It’s a problem that’s been around that people have been trying to solve, but now that it can be solved, you’re seeing the heads of marketing and the heads of sales really excited about this category and buying this software to help their teams be more competitive,” Wahbe said. Wahbe named and as Highspot’s top competitors. The company’s differentiator is its sophisticated AI that helps identify what content should be surfaced at the right time. Wahbe got the idea for Highspot when he was working at Microsoft, where he spent 16 years equipping sales teams with necessary information to help craft perfect pitches to potential customers. He quickly realized it was a difficult task and made a bet that others were experiencing the same problem. He founded the company seven years ago with former colleagues with and . Highspot was recently named to list for 2018, one of just two Seattle companies to earn the honor — Outreach, another fast-growing sales tech startup and a , was the other. Seattle has established itself as a hub for enterprise software, led by giants such as Microsoft, homegrown startups, and satellite offices for big companies including Salesforce. Wahbe emphasized Highspot’s commitment to Seattle, saying he didn’t plan to expand its offices internationally anytime soon. “It’s a little bit against the grain, but we really think the best way to build great software is to be here in Seattle,” Wahbe said.
Pebblebee, makers of Finder smart tracking device, find a big investor as part of effort to raise $10M

Pebblebee, makers of Finder smart tracking device, find a big investor as part of effort to raise $10M

12:54pm, 2nd April, 2019
The PebbleBee BlackCard is a new credit-card-thin tracking device that can help locate a lost wallet or anything else. (Pebblebee Photo) , the Bellevue, Wash.-based startup that makes a smart tracker to help people find missing keys and more, has found an investor. The company, founded by engineers Daniel Daoura and Nick Pearson-Franks, has landed a “substantial investment” from , a division of the massive Japanese wireless carrier KDDI Corp. The amount, which was not disclosed on Tuesday, contributes to what Daoura called an “ongoing $10 million funding round.” Soracom is a global provider of smart IoT connectivity, offering cloud-native wireless service designed specifically for the needs of connected devices. The company previously invested $5 million in Seattle-based balena.io (). Pebblebee has been making moves toward growing its brand and reach since last fall when it landed the Finder tracking device across the U.S. Daoura told GeekWire that the product sold really well and “proved the market” and they have expanded to Canada and other countries. He said they started entertaining the idea of looking into acquiring capital because growing the consumer brand requires hefty investment. “We got quite a bit of interest from the Bay Area as well as international VCs. Not so much local just because the nature of us being hardware and not software focused,” said Daoura, the startup’s CEO. Where’s the kid?! Pebblebee’s new BlackCard shown being tucked into a child’s jacket. (Pebblebee Photo) In the meantime, as Pebblebee aims to attract even more investors, the company isn’t slowing on development, and is releasing a new product this week called BlackCard. “It’s essentially a wallet tracker; it’s very thin — as thin as two credit cards — and it’s rechargeable,” Daoura said. The BlackCard has a range up to 500 feet — “definitely more than any other tracker out there today” — and will hold a single charge up to five months, and it emits a very loud buzzing sound. The price will be $29.99. BlackCard will launch along with a new and improved Pebblebee website on Wednesday. With eight employees, Daoura credits Pebblebee’s small team for bringing a Kickstarter vision to reality. “Their level of commitment and perseverance has been integral to our success,” he said. Soracom Americas CEO Eugene Kawamoto said in a news release that his company is passionate about identifying and supporting companies like Pebblebee. “Pebblebee’s hardware expertise and impressive patent library are already advancing the state of the art in the crucial asset tracking category,” said Kawamoto, who will take a seat on the Pebblebee board of directors. “By providing both smart connectivity and strategic investments, Soracom helps to accelerate IoT development and create a more connected world.”
‘Huge awakening’ in data privacy drives big growth for Seattle startup Integris

‘Huge awakening’ in data privacy drives big growth for Seattle startup Integris

11:29pm, 1st April, 2019
Integris CEO Kristina Bergman. (Integris Photo). Back in 2016, a Seattle startup called Integris with a modest $3 million in funding and a vision to help companies manage customer data with integrity. Fast-forward to 2019, when privacy issues are making daily headlines as politicians seek to rein in Big Tech, and business is booming for Integris. In a little over two quarters, Integris more than tripled its team to 30 full-time employees. The startup opened a second office in Vancouver, B.C. and is working with a number of Fortune 500 companies to help them implement data protection and privacy standards. Integris’ growth is driven by new laws in the U.S. and Europe that seek to crack down on tech companies that handle consumer data. The European Union is spearheading the effort with its broad General Data Protection Regulation. In the U.S., federal regulation has been sluggish as states step in to implement their own laws. Last summer, to give consumers more control over their data and dozens of other states are considering similar laws. Related: “When we started three years ago, most people couldn’t spell GDPR … but fast forward a few years and privacy is in the headlines,” said Integris co-founder Kristina Bergman. “It’s front page news in all the major publications and so the biggest thing that we’ve seen is a huge awakening among people everywhere about the impacts of privacy, the importance of privacy, and we’ve seen a lot of market maturity happen over the last few years.” Ironic as it might sound, big tech companies are . Apple and Microsoft have been actively promoting themselves as the secure, privacy-sensitive foils to their younger tech industry peers. It’s catching on. In March, Facebook by doubling down on encrypted, ephemeral messaging. But there is a growing concern in the business community about a future in which companies that handle consumer data are forced to comply with different laws in every state. “The concern is that if the federal government doesn’t step up and unify it in the way that Europe unified privacy legislation under GDPR, we’re going to end up with a privacy legislation framework in the U.S. that’s incredibly fractured, very hard to comply with, and not really feasible and implementable,” said Bergman. That fear is leading a number of tech leaders to support a federal privacy law that would pre-empt state regulations. Related: Integris surveyed 258 business executives at companies with 500 employees or more and at least $25 million in annual revenue as part of released Monday. Of those surveyed, 80 percent believe there should be a federal privacy law, though they may not be ready for it. About half of the respondents said they take inventory of the personal data they store just once a year or in response to an audit. However, 88 percent said their companies are increasing their data privacy management budgets in 2019. “What’s been a boon to the business is not the murkiness but the opportunity that privacy presents,” Bergman said. “In our discussions with companies, they’re looking at privacy increasingly as a differentiator for their business … they look at that as an opportunity to differentiate against their competition by being able to prove that they’re operating with integrity, they’re treating customer data with the utmost care, and they can prove it.” Integris’ goal is to help companies set up best practices in data privacy. The company uses machine learning and other technology to map a company’s sensitive data, apply regulatory obligations, and automate actions like encryption and deletion. On top of its initial $3 million round, last summer to amp up its regulatory compliance services.
Seattle startup ioCurrents raises $5M to bring big data to the high seas

Seattle startup ioCurrents raises $5M to bring big data to the high seas

1:16pm, 26th March, 2019
(Bernard Spragg Photo via Flickr) For , maritime data is both personal and professional. King grew up sailing in New England and is now the owner of “Northern Lights,” a vintage Coronado 41 sloop that he restored. ioCurrents CEO Cosmo King. (ioCurrents Photo) He’s also the CEO and co-founder of Seattle startup , which today announced a $5 million investment to grow its platform for collecting and analyzing real-time data for the maritime industry. The company’s platform, MarineInsight, collects reams of data from various pieces of ship machinery and analyzes it in the cloud whenever a connection is available. The software then suggests actions based on any problems it finds or anticipates; it can help reduce fuel costs or prevent engine failures, for example. The startup has customers in commercial shipping, fishing and passenger industries. King was formerly an engineer at Isilon Systems, a Seattle startup that was acquired by EMC in 2010 for $2.25 billion. He launched ioCurrents in 2015 with co-founder and CTO. “This additional investment will allow ioCurrents to build on our existing success, and provide even more value to the maritime industry as a whole,” King said . The Series A round, which brings the company’s total amount raised to $6.4 million, was led by . Imagen, a Seattle-based venture capital firm focused on data and software startups, has also invested in Seattle-area companies such as and outdoors app maker BaseMap. “ioCurrents is defining itself as the market leader in the development of real-time, predictive analytics to the maritime industry,” John Polchin, managing director of Imagen, said in a statement. “Imagen’s investment will help ioCurrents capitalize on the global demand for their solutions and accelerate the company’s pace of product innovation.”
How earthquake patterns could let us know when the ‘Really Big One’ is coming

How earthquake patterns could let us know when the ‘Really Big One’ is coming

2:05pm, 16th February, 2019
A map of coastal Washington state and British Columbia shows the sweep of an episodic tremor and slow slip event, or ETS, from February to April 2017. The colors denote the time of the event as shown on the color-coded time bar at the bottom. The gray circles on the color bar indicate the number of tremor events per day. (UNAVCO Graphic / Kathleen Hodgkinson) WASHINGTON, D.C. — Is it the tick of Earth’s heartbeat, or a ticking time bomb? Either way, a 14-month pattern in seismic activity could serve as the start of a super-early warning system for the “Really Big One,” the massive earthquake that’s expected to hit the Pacific Northwest sometime in the next few centuries. The seismic ticks are known as episodic tremor and slow slip (“ETS”) events, and . They’re linked to the titanic clash between the Juan de Fuca tectonic plate and the North American plate, in a region known as the Cascadia subduction zone. The two plates grind into each other at a rate of an inch or two per year, about 25 miles below the surface. Usually, it’s a slow grind, but every so often, there’s a sharp spike in the rate of movement. Along the Washington state coast, the spike comes roughly every 14 months. (The most recent spike .) In California, the cycle takes 10 months. In Oregon, it’s more like 24 months. Based on historical and geological records, seismologists have determined that the Cascadia fault can produce catastrophic earthquakes, on the order of magnitude 9.0 or more. In 2015, worries about the potential effects of a big Cascadia quake led to an about the . , a geophysicist at Oregon State University, isn’t saying the Really Big One is coming anytime soon. But during a presentation at this week’s annual meeting of the American Association for the Advancement of Science, she said a steadily expanding network of seismometers and strainmeter could give us advance notice. The seismic detection network in the Pacific Northwest and California allows seismologists to map the pulls exerted by ETS events in three dimensions, day by day. “When there’s a little pull, it increases the risk, the stress increases, and the probability for a great earthquake increases,” Trehu said. “But it increases from one very small number to what’s still a very small number.” Trehu said the key thing to watch for is a quickening in the pattern of episodic tremors. “Potentially changes in the pattern, changes in that periodicity, could be indicative of something interesting,” she said. “But those are going to take longer monitoring times.” Efforts are already underway to extend the seismic monitoring network offshore through the , a project backed by EarthScope and the National Science Foundation with participation by the and the .
Xnor shrinks AI to fit on a solar-powered chip, opening up big frontiers on the edge

Xnor shrinks AI to fit on a solar-powered chip, opening up big frontiers on the edge

9:50am, 13th February, 2019
Xnor.ai machine learning engineer Hessam Bagherinezhad, hardware engineer Saman Naderiparizi and co-founder Ali Farhadi show off a chip that uses solar-powered AI. (GeekWire Photo / Alan Boyle) It was a big deal two and a half years ago when researchers the size of a candy bar — and now it’s an even bigger deal for Xnor.ai to re-engineer its artificial intelligence software to fit onto a solar-powered computer chip. “To us, this is as big as when somebody invented a light bulb,” Xnor.ai’s co-founder, Ali Farhadi, said at the company’s Seattle headquarters. Like the candy-bar-sized, Raspberry Pi-powered contraption, the camera-equipped chip flashes a signal when it sees a person standing in front of it. But the chip itself isn’t the point. The point is that Xnor.ai has figured out how to blend stand-alone, solar-powered hardware and edge-based AI to turn its vision of “artificial intelligence at your fingertips” into a reality. “This is a key technology milestone, not a product,” Farhadi explained. Shrinking the hardware and power requirements for AI software should expand the range of potential applications greatly, Farhadi said. “Our homes can be way smarter than they are today. Why? Because now we can have many of these devices deployed in our houses,” he said. “It doesn’t need to be a camera. We picked a camera because we wanted to show that the most expensive algorithms can run on this device. It might be audio. … It might be a way smarter smoke detector.” Outside the home, Farhadi can imagine putting AI chips on stoplights, to detect how busy an intersection is at a given time and direct the traffic flow accordingly. AI chips could be tethered to balloons or scattered in forests, to monitor wildlife or serve as an early warning system for wildfires. Xnor’s solar-powered AI chip is light enough to be lofted into the air on a balloon for aerial monitoring. In this image, the chip is highlighted by the lamp in the background. (Xnor. ai Photo) Sophie Lebrecht, Xnor.ai’s senior vice president of strategy and operations, said the chips might even be cheap enough, and smart enough, to drop into a wildfire or disaster zone and sense where there are people who need to be rescued. “That way, you’re only deploying resources in unsafe areas if you really have to,” she said. The key to the technology is reducing the required power so that it can be supplied by a solar cell that’s no bigger than a cocktail cracker. That required innovations in software as well as hardware. “We had to basically redo a lot of things,” machine learning engineer Hessam Bagherinezhad said. Xnor.ai’s head of hardware engineering, Saman Naderiparizi, worked with his colleagues to figure out a way to fit the software onto an FPGA chip that costs a mere $2, and he says it’s possible to drive the cost down to less than a dollar by going to ASIC chips. It only takes on the order of milliwatts of power to run the chip and its mini-camera, he told GeekWire. “With technology this low power, a device running on only a coin-cell battery could be always on, detecting things every second, running for 32 years,” Naderiparizi said in a news release. That means there’d be no need to connect AI chips to a power source, replace their batteries or recharge them. And the chips would be capable of running AI algorithms on standalone devices, rather than having to communicate constantly with giant data servers via the cloud. If the devices need to pass along bits of data, they could . That edge-computing approach is likely to reduce the strain of what could turn out to be billions of AI-enabled devices. “The carbon footprint of data centers running all of those algorithms is a key issue,” Farhadi said. “And with the way AI is progressing, it will be a disastrous issue pretty soon, if we don’t think about how we’re going to power our AI algorithms. Data centers, cloud-based solutions for edge-use cases are not actually efficient ways, but other than efficiency, it’s harming our planet in a dangerous way.” Farhadi argues that cloud-based AI can’t scale as easily as edge-based AI. “Imagine when I put a camera or sensor at every intersection of this city. There is no cloud that is going to handle all that bandwidth,” he said. “Even if there were, back-of-the-envelope calculations would show that my business will go bankrupt before it sees the light of day.” The edge approach also addresses what many might see as the biggest bugaboo about having billions of AI bugs out in the world: data privacy. “I don’t want to put a camera in my daughter’s bedroom if I know that the picture’s going to end up in the cloud,” Farhadi said. Xnor.ai was , or AI2, only a couple of years ago, and the venture is with millions of dollars of financial backing from Madrona Venture Group, AI2 and other investors. Farhadi has faith that the technology Xnor.ai is currently calling “solar-powered AI” will unlock still more commercial frontiers, but he can’t predict whether the first applications will pop up in the home, on the street or off the beaten track. “It will open up so many different things, the exact same thing when the light bulb was invented: No one knew what to do with it,” he said. “The technology’s out there, and we’ll figure out the exact products.”
Xnor shrinks AI to fit on a solar-powered chip, opening big frontiers on the edge

Xnor shrinks AI to fit on a solar-powered chip, opening big frontiers on the edge

9:20am, 13th February, 2019
Xnor.ai machine learning engineer Hessam Bagherinezhad, hardware engineer Saman Naderiparizi and co-founder Ali Farhadi show off a chip that can use solar-powered AI to detect people. (GeekWire Photo / Alan Boyle) It was a big deal two and a half years ago when researchers the size of a candy bar — and now it’s an even bigger deal for Xnor.ai to re-engineer its artificial intelligence software to fit onto a solar-powered computer chip. “To us, this is as big as when somebody invented a light bulb,” Xnor.ai’s co-founder, Ali Farhadi, said at the company’s Seattle headquarters. Like the candy-bar-sized, Raspberry Pi-powered contraption, the camera-equipped chip flashes a signal when it sees a person standing in front of it. But the chip itself isn’t the point. The point is that Xnor.ai has figured out how to blend stand-alone, solar-powered hardware and edge-based AI to turn its vision of “artificial intelligence at your fingertips” into a reality. “This is a key technology milestone, not a product,” Farhadi explained. Shrinking the hardware and power requirements for AI software should expand the range of potential applications greatly, Farhadi said. “Our homes can be way smarter than they are today. Why? Because now we can have many of these devices deployed in our houses,” he said. “It doesn’t need to be a camera. We picked a camera because we wanted to show that the most expensive algorithms can run on this device. It might be audio. … It might be a way smarter smoke detector.” Outside the home, Farhadi can imagine putting AI chips on stoplights, to detect how busy an intersection is at a given time and direct the traffic flow accordingly. AI chips could be tethered to balloons or scattered in forests, to monitor wildlife or serve as an early warning system for wildfires. Xnor’s solar-powered AI chip is light enough to be lofted into the air on a balloon for aerial monitoring. In this image, the chip is highlighted by the lamp in the background. (Xnor. ai Photo) Sophie Lebrecht, Xnor.ai’s senior vice president of strategy and operations, said the chips might even be cheap enough, and smart enough, to drop into a wildfire or disaster zone and sense where there are people who need to be rescued. “That way, you’re only deploying resources in unsafe areas if you really have to,” she said. The key to the technology is reducing the required power so that it can be supplied by a solar cell that’s no bigger than a cocktail cracker. That required innovations in software as well as hardware. “We had to basically redo a lot of things,” machine learning engineer Hessam Bagherinezhad said. Xnor.ai’s head of hardware engineering, Saman Naderiparizi, worked with his colleagues to figure out a way to fit the software onto an FPGA chip that costs a mere $2, and he says it’s possible to drive the cost down to less than a dollar by going to ASIC chips. It only takes on the order of milliwatts of power to run the chip and its mini-camera, he told GeekWire. “With technology this low power, a device running on only a coin-cell battery could be always on, detecting things every second, running for 32 years,” Naderiparizi said in a news release. That means there’d be no need to connect AI chips to a power source, replace their batteries or recharge them. And the chips would be capable of running AI algorithms on standalone devices, rather than having to communicate constantly with giant data servers via the cloud. If the devices need to pass along bits of data, they could . That edge-computing approach is likely to reduce the strain of what could turn out to be billions of AI-enabled devices. “The carbon footprint of data centers running all of those algorithms is a key issue,” Farhadi said. “And with the way AI is progressing, it will be a disastrous issue pretty soon, if we don’t think about how we’re going to power our AI algorithms. Data centers, cloud-based solutions for edge-use cases are not actually efficient ways, but other than efficiency, it’s harming our planet in a dangerous way.” Farhadi argues that cloud-based AI can’t scale as easily as edge-based AI. “Imagine when I put a camera or sensor at every intersection of this city. There is no cloud that is going to handle all that bandwidth,” he said. “Even if there were, back-of-the-envelope calculations would show that my business will go bankrupt before it sees the light of day.” The edge approach also addresses what many might see as the biggest bugaboo about having billions of AI bugs out in the world: data privacy. “I don’t want to put a camera in my daughter’s bedroom if I know that the picture’s going to end up in the cloud,” Farhadi said. Xnor.ai was , or AI2, only a couple of years ago, and the venture is with millions of dollars of financial backing from Madrona Venture Group, AI2 and other investors. Farhadi has faith that the technology Xnor.ai is currently calling “solar-powered AI” will unlock still more commercial frontiers, but he can’t predict whether the first applications will pop up in the home, on the street or off the beaten track. “It will open up so many different things, the exact same thing when the light bulb was invented: No one knew what to do with it,” he said. “The technology’s out there, and we’ll figure out the exact products.”