Google is reportedly aiming to create technology that can be used to automate existing logistical systems and supply chains.
Oink -- a user-review app created by Kevin Rose's startup lab Milk before Milk was acquired by Google and Rose later joined Google Ventures -- never made it past its six-month birthday, but today the name and domain at least are getting another lease of youthful life. Virtual Piggy, a payments service aimed at minors (COPPA-compliant, with parental controls and an educational bent) is today rebranding as Oink after buying the domain, trademarks and social marks from Milk (not Google) for $ 57,500.
TechCrunch » Social
If you’re anything like me, you want your mobile device to fetch exactly what you ask for without requiring much effort on your part at all. And though Siri and Google Voice Actions do pretty well when you give simple commands or ask simple questions like “What temperature is it?” or “ Why is Tom Daley trending on Twitter?” they are less likely to be able to come up with a list of your beloved’s favorite restaurants from the conversations you’ve had via email.
Today Microsoft flipped the switch on Student Advantage, a program, announced in October, that extends the availability of Office to students of educational institutions that pay for Office 365 for their staff and faculty.
According to Microsoft, 35,000 educational institutions are eligible for Student Advantage, which provides access to the ProPlus SKU of Office 365, again provided that its paid staff are current users of Office 365 ProPlus or Office Professional Plus.
Office 365 ProPlus includes Access and Lync, making it a robust set of tools. Microsoft took a dig at Google in its announcement, stating that “[e]ven Google's own job postings require competency with Microsoft Office tools.”
What this means in practice is that Microsoft is lowering the marginal cost of Office for students to zero, while guaranteeing itself revenue through contracts with universities and the like. Microsoft cannot afford to cede mind and market share to Google, which provides a free Office competitor, and it must preserve its revenue from the product, which is a key profit source.
Office 365 ProPlus generally costs around $ 12 per month, per user, so the amount of ‘free' software that Microsoft will provide is non-trivial. To protect Office from low, or zero-cost competitors, it's probably sensible for it to sacrifice some revenue opportunity to keep up its primacy in the productivity market.
Top Image Credit: Flickr
You can't teach an old phone new tricks.
Hah! Just kidding. Of course you can. This isn't 1998.
Agent is an app that aims to make your Android smartphone just a little bit smarter, using all of your phone's sensors to detect what you're up to and tweak your settings automatically. Driving? It'll automatically respond to texts to let people know you're busy, and remember where you parked your car. Sleeping? It'll only let the most important calls through.
Agent is a spin-off, of sorts, of another Egomotion product called "Trigger" (or, as it was once known, "NFC Task Launcher"). With Trigger, Egomotion sells packs of programmable NFC tags which can fire off actions on your phone. Want your phone to silence itself and set an alarm when you go to bed? You'd stick one of their NFC tags on your nightstand, then set up a series of tasks to fire whenever that tag is detected. Want it to automatically launch your favorite music app when you get in the car? Tuck one of the tags into your cup holder.
In time, however, the team realized that many of the most popular use cases didn't really require NFC. Instead of an NFC tag on your nightstand, why not just auto-silence the phone during certain hours? Instead of hiding a tag in a cupholder, why not just detect when the user is connected to their car's Bluetooth? Thus, Agent was born.
Agent takes the core concept of Trigger and boils it down to its essence. Gone is any mention of NFC tags, instead relying on the handset's built-in capabilities - things like its accelerometer, clock, or WiFi/Bluetooth. Gone is the relatively complicated task setup process, with Egomotion instead providing a small set of pre-built actions that they call "Agents".
At the moment, the app's got five different built-in agents:
- Battery Agent: When your battery starts to fade, the battery agent kicks in to irk a bit more life out of your phone. You can tell it to automatically dim your screen, turn off automatic data syncing, or turn off Bluetooth. Once you're plugged in, it'll automatically flip everything back on.
- Sleep Agent: Automatically silences your phone between specified hours, but with a clever white-listing system. You can specify which contacts are allowed to wake you, and allow for repeat (and thus likely urgent) calls to ring through. It can auto-reply to texts, telling the user to reply "urgent" if it's an emergency (at which point, your phone will ring loudly to wake you up). You can tell this agent to only start if your handset is plugged in. That way, it probably won't silence your phone during a night out at the club.
- Parking Agent: Attempts to automatically remember where you've parked your car. By default, it works by detecting your speed; once you've stopped moving over a certain speed for more than a few minutes, it figures that you've parked your car and marks the location accordingly. Of course, doing things like riding the BART might fire off a false positive, so you can tell the Agent to base its logic off Bluetooth connectivity if it's an option in your car.
- Meeting Agent: Silences your phone during meetings. Uses your Google Calendar to determine your meeting schedule.
- Drive Agent: Uses bluetooth to detect when you're in your car. Can automatically silence your phone, read your texts aloud, and respond to incoming texts to let them know you're driving.
The company says that they've got more agents in the works, potentially offering add-on agent "suites" tailored to certain use cases - one set that'd be good for school, one set that'd be good for work, etc. That way they can keep adding more functionality without complicating the core application.
If you're a battle-tested Android expert, Agent's tricks might not raise an eyebrow. "Pft, I've got Tasker!" you say. "And I rooted my imported HTC J One and flashed it with a custom rom that does all this ages ago."
For the less intense folks (read: most people) out there, though, Agent should hit a sweet spot. It's simple, it does exactly what it promises to do, and the setup is very straightforward and well thought out.
My one hesitation: while I normally hate when people say "But what if company X just decides to do this", it's a pretty valid concern here. With all of the data that Google gobbles up and pipes into Google Now, it's almost certain that they're tinkering away with similar concepts right this second.
Actually, it's not almost certain. It is certain. Google-owned Motorola has already released an app that they call Assist, which aims to do much of the same stuff that Agent does. As Egomotion co-founder Kulveer Taggar pointed out to me, Assist only works with a handful of Motorola phones, whereas Agent works on many, many Android phones. But Moto's handsets tend to be a test bed for Google (See: the always-listening "Okay Google" voice command debuting on the Moto X months before being integrated into Android 4.4). If the concept proves popular, how long will it be before Google starts tying such functionality right into the core of Android itself?
In the mean time, though, it seems like Egomotion is on to something: according to the company's stats, 95% of agents that get turned on, stay on.
The app, normally $ 1.99, is on sale for $ 0.99 for the Thanksgiving weekend.
Google is upgrading its Google Place for Business service so users can collect everything their customers are saying about them online in a single, viewable place.
Buffer, the social media scheduling service, experienced a potentially brand-damaging security breach last month that saw a slew of weight-loss spam posted to Twitter and Facebook on behalf of its users. And although it turned out to be the company's database provider, MongoHQ, that was the origin of the compromise, Buffer, by its own admission, was squarely to blame as it hadn't encrypted access tokens for the social media services it supports. Today the startup is announcing a host of new security measures, including encrypting user email addresses and access tokens, and 2-step login, in a bid to restore confidence in its wares.
“With all that trust given to us, despite the big mess, we wanted to really step up our game in terms of safety and security,” writes the company on its blog. “For the past few weeks, we have been focusing on making Buffer the safest, most secure way for you to manage and publish to your social media accounts. We have a number of awesome things to show you. The most important step in this process is a feature we're announcing today: 2-Step Login”.
An optional setting for all Buffer users, 2-step login requires an additional security code to login to your Buffer account. It can be powered by SMS, requiring each user to register a mobile phone number with their account, or Google Authenticator, Google's own 2-step authentication app for iOS and Android. Both single and team Buffer accounts are supported. It's worth pointing out, however, that 2-step login, though a very sensible move, especially since a brand's image is essentially in the hands of its employees and Buffer's security, wouldn't likely have prevented the recent breach. But it's a good thing, nonetheless.
To that end, Buffer has also added encryption for user emails addresses and access tokens for social media services (Twitter, Facebook, Google+ etc.), preventing exactly the kind of hack that caused last month's “big mess”, as well as having all of its own team members change their passwords and set up two-factor authentication (where possible) on accounts for Google, Github, Stripe, HipChat and Dropbox. The thing about security is it's only as strong as the weakest point in the chain, and that is more often people.
TechCrunch » Social
Dianping, China’s Largest Restaurant Review Site, Turned Down An Acquisition Offer From Google China
Dianping, the restaurant review platform that is sometimes referred to as the “Yelp of China,” turned down an acquisition offer from Google China in early 2007, co-founder Edward Long revealed at the TechCrunch/Technode event in Shanghai. Google China valued the company, which was then four years old, at less than $ 100 million. After Dianping refused its buyout offer, Google China made a Series B investment in the site of a few million dollars. Dianping's other investors include Trust Bridge Partners, Sequoia Capital, QiMing Ventures and Lightspeed Venture Partners and it has raised almost $ 180 million in funding.
Long said Google China licensed its content for local search products. But then in 2010, Google pulled its search engine out of China, which Long says was a loss for Chinese consumers.
“If Google had stayed, Baidu wouldn't have become so strong and it wouldn't be able to charge so much for online advertising,” Long says, referring to China's largest online search company, which now holds a 63% share of the market according to analytics firm CNZZ.
He added that large international companies like Google and Yahoo have had a hard time gaining a foothold in China because their massive size prevents them from localizing quickly enough.
Rumors that Dianping is gearing up for an exit in the near future are untrue, Long says. The ten-year-old restaurant review platform, which has more than 75 million monthly active users and 6 million merchants in 2,300 cities throughout Asia, recently denied reports that it is being courted by Baidu with an acquisition offer of $ 2 billion. Dianping CEO has said that the company wants to go public in five years with an estimated valuation of more than $ 10 billion, but Long refused to confirm that amount at the event.
In an interview after his panel, Long explained that he feels Internet companies shouldn't go public until their market becomes more stable, because if a company issues too many quarterly forecasts that don't pan out, it will quickly lose the faith of investors.
Dianping monetizes through a combination review and e-commerce model, selling membership cards and e-coupons, and analytics and data for restaurant owners. It plans to open a marketplace for offline services soon.
Image source: Dianping.com
Apparently working at Google isn't as heavenly as it's cracked up to be.
okia is preparing to become a very different animal. The company whose name is still synonymous with mobile phones in certain parts of the world will - barring a last-minute bout of nerves from the company board - hand off its devices & services unit to Microsoft early next year in exchange for $ 7.2 billion. Which means that in 2014 Nokia is going to have both the time and the money to refocus its efforts elsewhere - and one key area for the future of the company, indeed one of only three remaining business units at Nokia, is maps & location services, under its HERE brand.
No surprise, then, that Nokia is revving up its engines in the location space. Notably, it's fully integrated the digital mapping giant Navteq - which it acquired way back in 2007, but was running as an independent business up until last year.
On its earnings conference call last month it also talked up the prospects for HERE to compete with Google Maps, and win new business in consumer facing use-cases such as on mobile devices. It's also revving HERE's engines literally, with a plan to triple the number of cars it has wearing down rubber mapping new roads and places, and doing the ongoing map maintenance required to keep the data fresh.
Nokia's HERE business sub-divides into three main divisions - its consumer-facing offering (of which Microsoft is among its top three customers); a b2b/enterprise play, where it works with customers such as Oracle, by, for instance, helping them track their flow of goods to better understand their own businesses; and a very big automotive business, built upon its Navteq acquisition and its turn-by-turn navigation specialism.
The automotive segment is currently the largest of HERE's divisions - as you'd expect, given how much Nokia spent on buying its way into the market when it purchased Navteq, for $ 8.1 billion. Globally, four out of five cars with dashboard navigation are using Nokia's maps. (“Basically every major car company is working with us, at some level,” it says.)
Across all HERE's division, Nokia's maps licensing and location data platform business has “hundreds” of customers.
From static to suggestive
This week Nokia was showing off one of its new-look HERE cars, pictured above, to U.K. journalists, giving live demos of the street-level mapping technology. The flagship tech on show here is the 3D mapping technology it acquired from Earthmine this time last year. Nokia is using this to compete with Google Street View in urban areas but also, more broadly, as the foundation for evolving HERE beyond the legacy Navteq specialism of turn-by-turn navigation.
Building more dynamic and interactive 3D maps is where Nokia sees the future for HERE, says Stuart Ryan, director of Maps and Everyday Mobility at Nokia. A smart, personalised, context-aware location cloud platform may also represent the future for Nokia's business, post-mobile phones.
It's about trying to reinvent and redefine the whole concept of what it is to see and visualise and interact with a map.
What kind of maps is it aiming to build? “Immersive, living, rich” maps that draw on multiple content sources and applications and change based on context, or display data in new and playful ways, says Ryan. Maps that also ”know and understand you, and suggest things for you”.
“Maps behave, react, inform in a way that's very, very relevant and personal to you,” he tells TechCrunch. ”The challenge our boss gives us is redesign the maps experience for consumers.”
One example he brings up to illustrate Nokia's thinking here is map of a city that's transformed from being a realistic digital representation into an abstracted data visualisation that emphasises certain characteristics of the city's makeup, and/or creates something that looks visually arresting, and is even purposefully un-map-like.
Another scenario he talks about is when visitors to a city want to know where local equivalents for places in their home town can be found - so, in other words, they're after a map that can show them where “the uptown version of Chicago in Berlin is”, for example.
“It's got to go from being a very largely static experience, granted built on turn-by-turn navigation, into a much more interactive and suggestive type of experience,” he says. “That's where it's going.”
Nokia's Earthmine vs Google Street View
It was too dark in London's low-light winter environs to take the HERE car for a spin when TechCrunch checked it out, but the driver showed the system working in the garage. Nokia's Ryan was also on hand to answer questions.
Nokia's HERE fleet is not doing all the mapping itself. The company has multiple sources for its maps, and even obtains some mapping data via community (crowdsourcing) contributions to augment the rubber on the tarmac approach, although Ryan stressed that it has teams checking mapping data submissions to ensure their quality, where possible. It also draws in other third party sources, such as data from government agencies.
The HERE car driver showing off the kit inside the demo car had recently been mapping the North East of England, and said he's typically out from 7:45 a.m. to 3:45 p.m.-4:45 p.m., depending on weather and light conditions. He typically covers around 200 miles a day.
The hard drive in the car - which was sited on the front seat, inside the processing unit - holds a terabyte of data. After around a week's worth of driving 64% of it had been filled.
The HERE car's modular rig, which is mounted on the roof, contains a LiDar sensor which captures the 3D data - grabbing objects and textures as the car moves along the road - by sending out signals that bounce back to build up a “3D point cloud”.
“This is what creates a 3D digital representation of the world around us,” says Ryan. “It's all managed as an X, Y, Z point so we have the location and height information… It's similar to laser and sonar, those kinds of technologies. It's based on light pulses and then when it comes back in we just store the location. And we create a composite view, then, of the real-world in a 3D construct. “
Here's an example of the street level imagery being captured by Nokia's rig:
And while, on the surface, Nokia's street level view may look much like Google's Street View, Nokia argues that because its technology captures 3D data rather than just taking static photographs, it offers a superior foundation for transforming the basic digital map canvas - by allowing for both new types of content and new ways of interacting with maps to be introduced.
“A static image product has a lot of limitations in terms of the utility it provides,” says Ryan when asked how Nokia's tech stacks up against Street View. He also says the imagery captured by Nokia's cameras is “higher accuracy, higher resolution, higher quality” than Google's Street View cameras, adding: “Which is part of the reason why we invested in the Earthmine technology”.
What he's not so quick to point out is that Nokia's street level mapping coverage is far less comprehensive than Google's Street View - Nokia is focusing street level mapping on metropolitan and urban areas, whereas Google, with five+ years of rubber in the game already, generally prefers to cover whole countries, as the below coverage map illustrates:
So when Nokia says it has mapped close to 200 countries, it's talking specifically about road-level mapping, not the street level 3D data. It's unclear exactly how extensive Nokia's street level mapping is at this point - Nokia has previously said the HERE cars will be visiting 27 countries this year – but even in Europe and North America where Nokia kicked off its 3D mapping effort coverage is selective, targeted on urban centres.
Nokia justifies that selective coverage by arguing that street level mapping is more relevant to urban areas where population density is highest - so, in the short term at least, HERE's Earthmine technology looks set to snub the roads less well-travelled.
But if you can't compete on quantity vs your biggest rival, there's always quality. And/or doing something new and different.
Maps that push and pull
“Our team is looking at the utility around how do you make the map more living, more real, more relevant to people - in terms of being able to provide more detail, and make the panoramas around you and buildings around you more relevant in terms of if you think about orientation and guidance,” says Ryan.
“But also… this [Earthmine technology] will allow us to make the map in and of itself more of a canvas and an interface - so we want to get away from the legacy of put something in a search box or a text box and actually scroll the map, select objects on the map, make them reveal things to you and go from there.
“Because everything we have is actually a data model, and not a static image, that allows us to do an awful lot of these use-cases.”
Introducing new “use-cases” to its maps is on the cards for Nokia “later next year”, according to Ryan. Fleshing out those use-cases a little further he lists the following: “truer 3D rendering of what's around you; making things much more interactive; allowing for certain things to render in certain ways - they could be very, very detailed, or they could be a little bit more occluded”.
Another feature he discusses is transitions - i.e. being able to move map users through a 3D view that takes in aerial imagery and moves down through different levels of detailed 3D content, before they arrive at street level. The aim being to wow users, with both the design and the feel of moving around the map.
He also talks about allowing users to explore cities via maps in a more tactile way using a 3D mapping canvas that lets them touch a building to see what it is, rather than having to type something into a text box or click around on rails. “You don't click an action to route to… you just click on the building itself to route to,” he explains.
The HERE car's modular mapping rig could also potentially be attached to other types of vehicles, according to Ryan - even something that's airborne. “If there's a way that we can find to mount this technology on the right kind of vehicle, can we also capture aerial imagery? Can we do other things as well? So it's extensible in terms of how we can use it.
“Our view is on this you differentiate on this more through actually making richer, deeper, interactive, immersive experiences with the map and really evolve how people think about a 2D or a 3D map. That's the aspiration for us,” he continues.
“It's about trying to reinvent and redefine the whole concept of what it is to see and visualise a map, and interact with a map. That's what no one has solved so far.”
From HERE Cars To Driverless Cars
As well as the LiDar sensor, the rig on Nokia's HERE cars carries a GPS sensor, and a Street View-esque camera that snaps the view from four sides so it can piece the visual scenery back together with the 3D mesh to colour in that data.
The other sensor on board the rig is called an IMU - aka an inertia measurement unit - which captures detailed road characteristics, such as the degree of incline or bend, which Nokia then sells back to customers in the automotive sector to improve their own in-car systems.
“This is a very automotive-centric sensor,” says Ryan of the IMU. “So it's got an accelerometer, gyroscope, it helps understand direction of travel, velocity of the vehicle, but also very, very important for understanding things like slope of a road, curve information, gradient, for products and use cases around ADAS: advanced driver assistance systems.
“So when you think about things like lane departure warnings, adaptive cruise control, curve warnings and use cases around fuel efficiency and route optimisation - and in essence building the foundation for how we think about automated or autonomous driving, that's one key thing that a sensor like this contributes.”
Nokia announced a partnership with Mercedes-Benz back in September to work on developing driverless car technology, noting at the time:
The requirements put on maps by autonomous driving go beyond what has been available until now. The very exact precision of lane width, road sign locations and other road network attributes are all crucial components needed to provide routing for autonomous vehicles. This is an area Mercedes-Benz and HERE will continue to explore together, as automakers and HERE spearhead the innovation for future commercial autonomous vehicles.
There's no time-frame on when such a car could make it to market - and in any case, the technical challenges of building self-driving cars are matched by huge legal and regulatory hurdles that stand just as tall, if not even taller barriers to commercialisation. So developing guidance systems to help steer self-driving cars is definitely a long-term play for Nokia; the road HERE's automotive business unit may ultimately end up travelling along one day.
In the meantime Nokia sees plenty of areas where it can more gradually extend the HERE platform's capabilities in the automotive sector - so that it gets even smarter about assisting drivers by collating and redistributing more relevant real-time data.
“There's a number of projects and discussions we're having with major OEMs, major car companies around this,” says Ryan. “A lot is happening in that area. Basically as all of these sensors [on the HERE car], plus potentially other sensors within the vehicle, continue to tag and have user information we can use that to feed back to a platform or service… and be able to aggregate that and deliver that back in the form of some value-add service to people.
“One example… [might be] if you're on the motorway and a car is 25 miles ahead but all of a sudden its windshield wipers start going, if you could signal that back to a platform, you could signal back to the car behind you that it's raining ahead. So all kinds of things.”
Of course, that example would require car makers to add the necessary sensors to windscreen wipers. But Nokia has various other data sources feeding into HERE's network already - such as traffic data from enterprise customers with large delivery fleets, and anonymised data gathered from its own maps and mobile users - which it refers to as “probe data”. (You can bet the deal with Microsoft to buy Nokia's devices & services division requires that this pipeline of probes remains open to HERE.)
Maps made of multiple data sources
One area Ryan mentions where probe data could be used to augment Nokia's maps with useful, contextual information is by gathering real-time transit data - i.e. where public transport networks don't make that data openly available.
“We can build estimated schedules based on the lines, the length of the journey between stops, the speed of the train and then we get timetable data,” he says. “The next thing people want to understand is interrupted service, delays, real-time information for transit as well so getting user feedback for that kind of content is interesting.”
Probe data also feeds back into the HERE business in another way: to help Nokia identify when a new road has been built that it needs to send a car out to map. Or to build “heat maps” of activity based on where people are doing map searches and how they are using other Nokia apps (some of which are available on iOS and Android, expanding Nokia's reach beyond its own legacy/Microsoft's mobile platforms).
You could call it Nokia's very own inner Foursquare.
“We actually use [probe data] on our map assets. You can see how we think about activity heat-maps - where the most interesting places and restaurants and POIs are in a city or an area are,” says Ryan.
Probe data, third party data sources, community-contributions and Nokia's own Earthmine-equipped cars all feed their intel into the HERE platform - Nokia says it uses 80,000 different sources to layer content into its maps. The blood, if you like, that is being pumped through HERE's digital veins to animate increasingly dynamic maps.
How competitive that data set is vs the Mountain View-sized hoard that Google's mapping business rests on top of is of course a key question.
Looking at the competitive landscape
Nokia says it is planning to have a fleet of “hundreds” of HERE cars on the road by the end of 2014. It won't be more specific about exactly how many cars it has, driving around hoovering up the topology. Google also won't put a figure on its Street View fleet, saying only that it's driven more than five million miles of unique road since kicking off the project in 2007.
As with Google's fleet, Nokia's HERE cars are driving around all the time - a necessity to keep its maps updated, and also to expand street level mapping coverage to better compete with Google.
This never-ending process is clearly a huge cost base for HERE, along with the unit's 6,000-strong headcount. Nokia's operating expenses for HERE were €153 million in Q3 this year, which is small relative to other divisions of the business but bear in mind that those divisions are involved in manufacturing hardware vs HERE being a software-focused unit.
There's a lot of blood in the water with Google around, and other players as well, but there's huge upside in terms of where the location area is going - and nobody's really solved it yet.
Focusing HERE's street level mapping coverage on cities/towns is one way Nokia can cap the costs associated with gathering the more sophisticated 3D data - while being able to innovate on the technology and software side to produce more dynamic and interactive maps for those urban centres in future. And, so Nokia hopes, steal some of Street View's thunder.
In the automotive sector, Nokia primarily competes with TomTom right now, although Ryan concedes that Google is looking to get a “stronger footing” in that industry - so there's certainly a sense that the company is checking itself about getting complacent. And that's not surprising, given the painful lessons of its recent history. Nokia's mobile business went from being king of the world circa 2007 to also-ran a handful of years later, having watched Google's Android platform spread like a pandemic from a standing start in 2008 to the circa 80% marketshare of today.
Ryan is also not complacent when it comes to location startups - name-checking Waze as one example of a startup that managed to build significant momentum, with a crowdsourced traffic data platform, and which was of course ultimately acquired by Google.
“The opportunity for disruptors to come along is to solve that hidden use-case or that unmet need. The company that can combine a lot of what we do, and Google go, and Foursquare do, and Facebook do in a location cloud experience is the winning strategy, and that's why it's the strategy that we're pursuing,” he says.
That means Nokia's challenge, as a long-time player in the location space, is to take a step back from the turn-by-turn/navigation business it knows so well, through Navteq, and allow HERE to become more mobile and social, to be “more about solving more everyday life problems”, as Ryan puts it. “That is where there is an awful lot of untapped opportunity within maps and location.”
“We recognise there's a lot of blood in the water with Google around, and other players as well, but there's huge upside in terms of where the location area is going - and nobody's really solved it yet,” he adds.
With its phone-making days behind it, Nokia wants to plant its flag afresh on the location map - with a banner that reads not just HERE today; here to stay.
TechCrunch » Social