Frustrated Snap Social Influencers Leaving For Rival Platforms

Frustrated Snap Social Influencers Leaving For Rival Platforms

Sallia Goldstein

During a November 2014 trip to Los Angeles, Boston-based Snapchat creator Mike Platco stopped by Snap&;s Venice, CA offices to say hello. At the time Platco had amassed a sizable audience of followers — large enough that he was preparing to leave his day job to create content for Snapchat full-time. He figured a casual visit to Snap HQ, while perhaps unexpected, wasn&039;t untoward given his profile on the platform. But after trading messages with a few Snap employees, Platco was turned away at the door. There would be no visit.

“That was the very first in a long history of instances where Snapchat really solidified its position as giver of zero fucks,” Platco told BuzzFeed News. With approximately 500,000 Snapchat followers, he now works regularly with brands who pay him as much as $80,000 for a series of posts.

To paraphrase Patco, Snapchat hasn’t done much to foster a good relationship with creators — people who specialize in making the raw, goofy content that feels native to social media. These creators — particularly those who&039;ve emerged as social media personalities — can be quite powerful in aggregate. Many have passionate audiences willing to follow them from platform to platform. Now, after years of feeling neglected by Snapchat, some of these creators are fed up and taking their work and their fans elsewhere. That could be problematic for the newly-public company — just ask Vine, which last year suffered a catastrophic migration of top Viners to rival platforms like Instagram and, yes, Snapchat for similar reasons.

“Creators are starting to post less on Snapchat.”

“Creators are starting to post less on Snapchat,” said one manager of well-known social creators. “Where it used to be the primary platform they were making content for, now it’s like a secondary platform where they’ll make content for other platforms and repost it on Snapchat.”

Platco, who says he owes his fame to Snapchat, is now working to move his followers to Instagram. He feels it offers the type of support Snapchat hasn’t — listening to product feedback, providing post analytics and promotion. “I love Instagram,” Platco said. “Every single bad thing I could possibly say about Snapchat, I could say the opposite of how my relationship is going with Instagram.”

Platco spent the two weeks before Christmas pushing his Snapchat followers to his Instagram account with artistic posts and offers of “epic” Instagram-only giveaways. The campaign tripled Platco&039;s following on the photo-sharing service, though it’s still far smaller than the one he’s amassed on Snapchat. “I’m just going to keep grinding it out and come up with new and engaging ways to get my followers over to Instagram,” he said.

Instagram: @mplatco

Sallia Goldstein, another widely followed Snapchat creator, said she moved to Instagram two weeks ago — reluctantly. “It’s not because I necessarily want to move everything over to Instagram,” she told BuzzFeed News. “It’s because I have to.”

“It’s not because I want to move everything over to Instagram. It’s because I have to.”

Goldstein shifted her work to Instagram because Snapchat’s pokey performance on Android made the app effectively unusable for her. She was unable to work through those issues with Snap (messages to support went nowhere); So now Goldstein uses her Snapchat account to push her followers to Instagram. She’ll often start a story on Snapchat and tell her followers to head over to Instagram to watch the rest. “I would much rather make stories on Snapchat than Instagram. Snapchat has so many more creative tools than Instagram,” Goldstein said. But she&039;s had great success migrating followers from Snapchat to Instagram; Each promotion reliably sends between 500 and 1,000 followers across. Goldstein figures her Instagram following will match her Snapchat audience in a matter on months.

Like Goldstein, many of the creators BuzzFeed News spoke with are fond of Snapchat and the opportunity it’s provided them. “I think it’s a really great platform for being unique and creating content.” Shaun McBride, an outspoken Snapchat creator who goes by the name Shonduras, told BuzzFeed News. “I like Snapchat,” said Cyrene “CyreneQ” Quiamco, another creator. “I’m an artist and Snapchat is the paint.”

But these creators also dismayed by Snapchat’s seeming unwillingness to collaborate with them — specifically to refine its publishing tools to help them work more closely with the brands that pay them. For many creators who make their livings posting branded content to social media platforms, a toolset that helps them to measure and dissect the audiences they reach is crucial. According to the creators interviewed for this article, Snap doesn&039;t have that, nor is it particularly attentive to their requests that it develop one. “If Snapchat would just open up the lines of communication with creators and help us understand the platform better so we could work [more easily] with brands, we&039;d probably focus more on posting our best content on Snapchat,” said McBride.

youtube.com

So far that hasn&039;t happened. And creators are growing antsy; McBride has recently been focusing his attention on YouTube, where’s he’s built a following of nearly 1 million users. Meanwhile, Story views on Snapchat have declined precipitously, according to the social analytics company Delmondo. They slipped 40% from July 2016 to November 2016, a fall that began roughly with the introduction of Instagram Stories. “There is a very defined path to monetization for the aspiring online video star outside of Snapchat rather than on Snapchat,” Delmondo CEO Nick Cicero told BuzzFeed News.

Snapchat parent company Snap Inc. — which will begin trading publicly tomorrow at a heady initial share price of $17 — declined comment on record about creator claims of neglect. But longtime observers of the company say its lack of engagement with creators is strategic — a purposeful move to give Snapchat a friends and family feel that will differentiate it from rivals.

“This isn’t some mistake that Snapchat is not seeing,” Jason Stein, CEO of the advertising agency Laundry Service told BuzzFeed News. “They’re making a thoughtful decision to be something different than the other platforms.”

A Snapchat spokesperson confirmed that the company’s focus is on friends and family, not influencers. And that’s intentional.

Question is, can Snap continue to pursue that strategy and shoulder the associated risks with public market investors demanding that it live up to its sky-high valuation. You only need only look to Twitter for how a scenario like that can quickly turn ugly.

“While there are certainly cons to what they’re doing, there are also many pros,” Stein explained. “That’s what’s made them successful to this point. The question is how does that scale as a public company.”

Quelle: <a href="Frustrated Snap Social Influencers Leaving For Rival Platforms“>BuzzFeed

Oculus Just Slashed Its Prices, But VR Is Still Expensive

Oculus just cut the price of its Rift virtual reality headset from $599 to $499 and its controllers from $199 each to $99.

You can also buy a Rift and two controllers for a $598 bundle. The company, owned by Facebook, makes the Oculus Rift headset and the software that powers the Samsung Gear.

Oculus / Via oculus.com

The price cuts are meant to make the Rift more accessible to people turned off by its high cost. But $600 is still a big chunk of change. And to use an Oculus, you still need a computer with enough processing power and memory to download and run the games, and as mentioned on Oculus&; website, these kinds of computers often cost upwards of $1,000.

This doesn&039;t mean Oculus is failing.

Oculus won&039;t disclose how many Rifts it&039;s sold so far, but the New York Times reports that this kind of price cut isn&039;t indicative of struggling sales at Oculus. It&039;s more likely that, as the company smooths out its manufacturing and logistics, there are fewer recurring errors that bring up the average cost of a unit. The company wants to expand beyond its core audience of tech and gaming enthusiasts, so Oculus hopes the cuts will lure more people to virtual reality, according to the Times.

But its price isn&039;t helping.

Oculus has competition in the VR space. Sony recently announced it sold almost a million Playstation VR headsets just four months after the device&039;s debut in October 2016. Oculus has been on the market since March 2016. The Playstation VR, priced at $399, runs on the Playstation 4 console, also priced at $399.

Sony, which has sold 53 million Playstation 4 consoles, has an established lineage of dedicated partner studios and blockbuster titles. Oculus, much newer to the gaming industry, faces wariness about the return on investment for VR games. Oculus doesn&039;t have a game as big as Mario or Halo yet, though Oculus does plan to release one game per month from its internal studios in 2017.

In a blog post, Oculus executive Jason Rubin acknowledged that price seems be a determining factor in how well VR rigs sell. He notes that Playstation&039;s console headset is beating the Rift because of its competitive price, but that “even less expensive Mobile VR headsets, like our Gear VR device, are outselling Console VR.”

It&039;s true&; VR devices are mad expensive&033;

Giphy

The other powerful VR rig, the HTC Vive, will cost you $800 for a headset, two controllers, and two motion-sensing towers. And that price doesn&039;t even take into account the powerful PC you need tether it to. Vive said in a statement that it won&039;t match the Rift&039;s price.

Even Google Cardboard, just $15, relies on a smartphone with a data plan. Google also recently released the Google Daydream VR headset ($80), which currently only works with Google&039;s Pixel smartphone ($649).

So for some people, Oculus&039; price cuts still aren&039;t enough.

Quelle: <a href="Oculus Just Slashed Its Prices, But VR Is Still Expensive“>BuzzFeed

Snapchat's $24 Billion Valuation Sets A High Bar For Its Future

Snapchat CEO Evan Spiegel

Michael Kovac / Getty Images

Snapchat&;s parent company is expected to be valued at about $24 billion in its IPO, setting an high bar bar for a company that lost $515 million last year.

Investors will buy the company&039;s stock for an initial price of $17 a share, according to reports, creating expectations for dramatic growth in users and revenue at the company, which will spend the coming years trying to live up to those expectations.

Snap Inc, Snapchat&039;s parent company, brought in $405 million in revenue in 2016, while spending $925 million.

The $24 billion valuation is slightly above the range the company aimed for two weeks ago when its IPO plans were made public. It&039;s not uncommon for the price of shares in an IPO to rise slightly from the company&039;s initial estimate. About $3.4 billion worth of Snap shares will be sold, with existing investors taking home about $935 million from the sale and the company pocketing the rest. The sale will be the biggest tech IPO since the Chinese e-commerce giant Alibaba raised $22 billion in 2014.

At $24 billion, Snap is still considerably less valuable than the $104 billion Facebook was valued at when it went public in 2012, but more than twice Twitter&039;s current $11 billion valuation.

And speaking of Snapchat&039;s 140-character-cousin: Twitter has become a cautionary tale for what happens when a hot social media company goes public but can&039;t meaningfully grow its user base or live up to a sky-high valuation (Twitter&039;s market value hit almost $25 billion on its first day of trading). Some analysts and investors fear that the same fate could befall Snapchat, whose user growth has slowed down recently.

The company initially targeted a valuation in its IPO between $20 and $25 billion, the Wall Street Journal reported. At the end of 2016, the company calculated that the fair value of its shares was $16.33, according to its IPO filing.

Quelle: <a href="Snapchat&039;s Billion Valuation Sets A High Bar For Its Future“>BuzzFeed

Twitter Will Start Using Algorithms To Crack Down On Abusive Accounts

Ariel Davis/BuzzFeed News / Via buzzfeed.com

Today, Twitter will start relying on algorithms to identify and restrict accounts for engaging in “abusive behavior,” which the company defines as either “repeatedly Tweeting without solicitation at non-followers” or violating the Twitter terms of service. The change goes beyond what some thought would be simple keyword policing (read: swear words) by also considering the relationships between users when determining abuse.

Penalties may include making an account&;s tweets only visible to its followers, being forced to verify a phone or email address associated with the account, or being suspended for 12 hours or more. In a statement about the changes, the company&039;s vice president of engineering Ed Ho wrote, “Our platform supports the freedom to share any viewpoint, but if an account continues to repeatedly violate the Twitter Rules, we will consider taking further action.”

Twitter seems to expect this approach will have hiccups, as Ho acknowledged: “Since these tools are new we will sometimes make mistakes, but know that we are actively working to improve and iterate on them every day.” There isn&039;t a process to appeal any of the the penalties yet, though Twitter&039;s plan to “iterate every day” indicates that may change.

Twitter will also allow you to filter out notifications from accounts that do not have a profile photo or that list an unverified email addresses or phone number, which are sometimes signs that an account was created specifically to abuse others anonymously. You&039;ll also be able to decide how long you want to mute accounts, conversations, and keywords. These features resemble the quality filter and notifications settings that verified users have had for some time now.

Aside from its new filter options and algorithmic abuse policing, Twitter will start sending notifications about the status of reported tweets and accounts to the people who flagged them. Specifically, the company will ping you when it has received your report and when it decides to take action. Twitter previously didn&039;t notify users when it received harassment reports or when it made a decision on how to handle them, which left many in the dark about whether they had even successfully communicated with the company.

The social network has shipped a number of product updates in the past month, many of which were tools intended to combat harassment: hiding tweets and conversations that mention you if you&039;ve been blocked by the writer, a setting to only see replies by people you follow, hiding sensitive content in search, and the ill-fated hiding of notifications when you&039;re added to a list.

Twitter has struggled with online abuse and harassment since its inception, especially with the question of how to define abuse on it platform — something that may have hindered its recent failed attempt to sell itself.

Quelle: <a href="Twitter Will Start Using Algorithms To Crack Down On Abusive Accounts“>BuzzFeed

Facebook’s AI Is Better Than Humans At Picking Out Suicidal Posts

Facebook

Facebook is bringing its artificial intelligence expertise to bear on suicide prevention, an issue that&;s been top of mind for its CEO Mark Zuckerberg following a series of suicides live-streamed via the company’s Facebook Live video service in recent months.

“It&039;s hard to be running this company and feel like, okay, well we didn&039;t do anything because no one reported it to us,” Zuckerberg told BuzzFeed News in an interview last month. “You want to go build the technology that enables the friends and people in the community to go reach out and help in examples like that.”

Today, Facebook is introducing an important piece of that technology — a suicide prevention feature that uses artificial intelligence to identify posts indicating suicidal or harmful thoughts. The AI scans the posts and their associated comments, compares them to others that merited intervention and, in some cases, passes them along to its community team for review. The company plans to proactively reach out to users it believes are at risk, showing them a screen with suicide-prevention resources including options to contact a helpline or reach out to a friend.

“The AI is actually more accurate than the reports that we get from people that are flagged as suicide and self injury,” Facebook product manager Vanessa Callison-Burchold told BuzzFeed News in an interview. “The people who have posted that content [that AI reports] are more likely to be sent resources of support versus people reporting to us.”

Facebook’s AI will directly alert members of the company’s community team only in situations that are “very likely to be urgent,” Callison-Burchold said. Facebook says a more typical scenario is one in which the AI works in the background making a self-harm reporting option more prominent to friends of a person in need.

While suicide prevention is a new and unproven application for artificial intelligence, Dr. John Draper, project director for the National Suicide Prevention Lifeline — a Facebook partner, says it&039;s promising. “If a person is in the process of hurting themselves and this is a way to get to them faster, all the better,” he told BuzzFeed News. “In suicide prevention, sometimes timing is everything.”

In addition to AI monitoring for indications of self-harm, Facebook is rolling out a few other suicide prevent features as well. It will make a number of suicide-prevention organizations available for chats via Messenger and it will present suicide-prevention resources to Facebook Live broadcasters who are determined to be at risk. With a tap, a broadcaster can access suicide-prevention resources, visible on on their screen.

Facebook&039;s decision to maintain the live broadcast of someone who&039;s been reported as at-risk for self harm is clearly fraught. But the company appears ready to risk broadcasting a suicide if doing so gives friends and family members a chance to intervene and help. “There is this opportunity here for people to reach out and provide support for that person they’re seeing, and for that person who is using live to receive this support from their family and friends who may be watching,” said Facebook researcher Jennifer Guadagno. “In this way, Live becomes a lifeline.”

The debut of Facebook&039;s new suicide prevention tools come amid growing concerns about the company&039;s influence, and could raise concerns about digital privacy. But for Facebook, which has been working hard to take a more direct role in stopping suicide, AI could be a big step forward in evaluating potentially suicidal content and and making it easier for people to help friends in need. As Zuckerberg wrote in his February 16 letter, “Looking ahead, one of our greatest opportunities to keep people safe is building artificial intelligence to understand more quickly and accurately what is happening across our community.”

Quelle: <a href="Facebook’s AI Is Better Than Humans At Picking Out Suicidal Posts“>BuzzFeed

Angry Neighbors Protest Outside Snap Offices Ahead Of Highly Anticipated IPO

Gavin Stenhouse

As Snap Inc. gears up for its forthcoming IPO, some of the company&;s Venice, California neighbors rallied outside its doors, aiming to send a message to the fast-growing social media phenom: Get out.

On Tuesday afternoon, dozens of Venice locals gathered outside of Snap&039;s offices to protest what they say is an unwelcome transformation of a vital piece of Los Angeles.

“This is a public street and the community will not sit by quietly while Snap attempts to annex it for a private corporate campus,” 11 year Venice resident Laura Booth told BuzzFeed News.

Snap&039;s headquarters is scattered throughout multiple buildings in the quirky beachside enclave that&039;s home to surfers, eccentrics and now herds of tech employees. The company&039;s Venice footprint has ballooned ahead of an initial public offering expected to hit the market this later week. And, as BuzzFeed News reported last week, that growth is causing serious tension with neighbors, some of whom say Snap is turning Venice into “a horrible business park.”

Instagram: @cjgronner

Reached for comment, Snap told BuzzFeed News that it is looking beyond Venice for future expansion.

“We’re very grateful to be a part of the Venice community and we are sorry for any strain that our growth has placed on those who live and work here,” a Snap Inc. spokesperson said.. “We&039;ve partnered closely with local schools and nonprofits to be a good neighbor and we&039;ve always tried to help our community feel safer in a neighborhood that is all too often the victim of violent crime. We recognize that we are no longer the small startup that we once were and we are necessarily concentrating our future growth outside of Venice.”

Laura Booth

Laura Booth

Laura Booth

Quelle: <a href="Angry Neighbors Protest Outside Snap Offices Ahead Of Highly Anticipated IPO“>BuzzFeed

Video Shows Uber CEO Travis Kalanick Arguing With Driver Over Fares

Video Shows Uber CEO Travis Kalanick Arguing With Driver Over Fares

Shu Zhang / Reuters

Uber’s public relations crisis continues apace with no apparent end in sight.

On Tuesday afternoon, Bloomberg published a video in which CEO Travis Kalanick aggressively argues with an Uber driver who claimed he is earning less money after Uber cut fares. “Some people don&;t like to take responsibility for their own shit,” Kalanick exclaims, after his driver says he lost $97,000 because of Uber. “They blame everything in their life on somebody else. Good luck&;”

youtube.com

The publication of the dash-cam shot video is the latest in a parade of PR disasters for Uber. In January, Kalanick’s decision to sit on President Trump’s economic advisory group inspired a viral campaign in which the company saw about 200,000 users delete their accounts, according to the New York Times. Kalanick subsequently resigned from the council.

Then, in early February, a former Uber engineer penned a viral account of her experience at the company with detailed allegations of systemic sexism. In response, Uber launched an internal investigation into the accusations, led by former attorney general Eric Holder and Arianna Huffington, who sits on Uber’s board. A visibly emotional Kalanick apologized to his staff at an all-hands meeting and promised to “do better.”

Two days later, during a meeting with more than 100 women engineers, Kalanick was grilled about issues of sexism at Uber, according to an audio recording obtained by BuzzFeed News. “I want to root out the injustice,” he told those in attendance. “I want to get at the people who are making this place a bad place. And you have my commitment.”

Uber’s tensions with its drivers are well-documented. The company continues to grapple with lawsuits over the classification of drivers as independent contractors. Just last month, Uber paid the Federal Trade Commission $20 million to settle allegations that it advertised inflated estimates of how much its drivers earn on its website and in Craigslist job postings.

Kalanick’s video interaction with his Uber driver is in many ways a snapshot of those tensions — and one that Uber clearly did not expect to become public. Uber declined to comment on the video.

Uber says on its website that drivers are permitted by the company to record riders “for purposes of safety,” but notes that “local regulations may require individuals using recording equipment in vehicles to fully disclose to riders that they are being recorded in or around a vehicle and obtain consent.”

In California, a state with a two-party consent rule for recording confidential conversations, could the driver be in legal trouble?

“It was a risky move to publicize this video,” Eric Goldman, a law professor at Santa Clara University, told BuzzFeed News. “It’s unclear if the conversation between the Uber driver and the CEO would qualify as a confidential communication.”

Goldman said whether the conversation would qualify as confidential would depend on several factors, such as whether the dashcam was prominently visible, and whether for-hire vehicles could count as public spaces. Regardless of those questions, he said, lawsuits of this variety are uncommon and the optics around Uber suing one of its own drivers lower the odds of a lawsuit.

Said Goldman, “Uber’s CEO has much bigger problems in his life right now.”

Quelle: <a href="Video Shows Uber CEO Travis Kalanick Arguing With Driver Over Fares“>BuzzFeed

YouTube’s New App Lets You Watch Live TV

YouTube / Via tv.youtube.com

YouTube unveiled YouTube TV today, a standalone app that&;ll let you watch 40+ cable and broadcast channels via the internet for $35 per month. The service will launch in the spring at an unspecified date in “the largest US markets,” according to a YouTube statement. Key channels include ESPN, CBS, ABC, USA, FX, Fox News, E&;, the CW, and others. And just like a cable subscription, you can add premium channels like Showtime to your bundle for extra money per month.

The service resembles Dish’s Sling TV, Sony PlayStation Vue, and AT&T’s DirecTV Now, which allow people to watch live TV on traditional channels via the internet. Hulu is planning to release a similar service soon, according to the New York Post. Facebook has plans for a standalone TV app, and Apple, already a player with Apple TV, has announced plans for making original TV shows.

YouTube TV is separate from YouTube Red, the site’s premium content channel that requires a subscription, though subscribing to YouTube TV also gives you access to YouTube Red Originals. (Disclosure: YouTube Red has purchased web series from BuzzFeed). YouTube TV will be a standalone app downloaded to phones (both iOS and Android), tablets (same), or computers. In its announcement blog post, the company highlighted the ability to watch YouTube TV on traditional sets via the company’s Chromecast device.

You’ll be able to record live shows and save them to the app without storage limits, where you can keep them for up to nine months. Each subscription comes with the ability to create six personalized accounts and watch three concurrent streams at once. Recode reports that Google’s artificial intelligence software will power the service’s recommendation system. The company didn’t say how regular YouTube videos will interact with YouTube TV, but it is worth noting that TV will be a separate app from YouTube’s flagship downloadable service.

Justin Connolly, an executive vice president at Disney and ESPN, said in a statement that the service would allow networks to reach “young, mobile-first audiences.”

Quelle: <a href="YouTube’s New App Lets You Watch Live TV“>BuzzFeed

Optimizing rolling feature engineering for time series data

In this blog post, I want to talk about how data scientists can efficiently perform certain types of feature engineering at scale. Before we dive into sample code, I will briefly set the context of how telemetry data gets generated and why businesses are interested in using such data.

To get started, we know that these days machines are instrumented with multiple in-built sensors to record various measurements while it is in operation. Thus, these machines end up generating a lot of telemetry data that can be used once this data is transferred off these machines and stored in a centralized repository. Businesses these days hope to use their amassed data to help answer questions like, “When is a machine likely to fail?” or, “When does a spare part for a machine need to be re-ordered?” Eventually this could help them reduce time and costs incurred in adhoc maintenance activities.

After having built many models, I have noticed that typical telemetry data that gets generated from the various sensors in their raw format add very little value. Sensors by design can generate data at a regular time interval, thus the data consists of multiple time series which can be sorted by time for each machine to build meaningful additional features. So, data scientists, like me, end up enhancing the dataset by performing additional feature engineering on this raw sensor data.

The most common features I begin with are to build out rolling aggregates using my preferred statistical programming language on a sample dataset. Here are some code snippets on how I would generate rolling aggregates for a specific window size using R/Python for machines which records voltage, rotation, pressure, and vibration measurements by date. These code snippets can be run on any other local R/Python IDE, within a Jupyter notebook or within an Azure ML Studio environment.

R

Python

telemetrymean <- telemetry %>%
    arrange(machineID, datetime) %>%
    group_by(machineID) %>%

    mutate(voltmean = rollapply(volt, width = 3, FUN = mean, align = “right”, fill = NA, by = 3),
                  rotatemean = rollapply(rotate, width = 3, FUN = mean, align = “right”, fill = NA, by = 3),
                  pressuremean = rollapply(pressure, width = 3, FUN = mean, align = “right”, fill = NA, by = 3),
                  vibrationmean = rollapply(vibration, width = 3, FUN = mean, align = “right”, fill = NA, by = 3)) %>%
    select(datetime, machineID, voltmean, rotatemean, pressuremean, vibrationmean) %>%
    filter(!is.na(voltmean)) %>%
    ungroup()

temp = []
fields = [&;volt&039;, &039;rotate&039;, &039;pressure&039;, &039;vibration&039;]
for col in fields:
    temp.append(pd.pivot_table(telemetry,
                               index=&039;datetime&039;,
                               columns=&039;machineID&039;,
                               values=col).resample(&039;3H&039;, closed=&039;left&039;, label=&039;right&039;, how=&039;mean&039;).unstack())
telemetry_mean_3h = pd.concat(temp, axis=1)
telemetry_mean_3h.columns = [i + &039;mean_3h&039; for i in fields]
telemetry_mean_3h.reset_index(inplace=True)

For more details on a description of the end to end use case please review the R code and Python code.

Once my R/Python code is tested in the local environment with a small dataset and deemed fit, I would then need to move it into a production environment. I would now need to also consider the various options on how to scale the same computation for a much larger dataset while ensuring efficiency. I have noticed that it is often more efficient to work with data that is indexed for such large-scale computations using some form of SQL query. Here is how I translated the code originally written in R/Python into SQL query language. 

Sample SQL code

select rt.datetime, rt.machineID, rt.voltmean, rt.rotatemean, rt.pressuremean, rt.vibrationmean
from
(select avg(volt) over(partition by machineID order by machineID, datetime rows 2 preceding) as voltmean,
        avg(rotate) over(partition by machineID order by machineID, datetime rows 2 preceding) as rotatemean,
        avg(pressure) over(partition by machineID order by machineID, datetime rows 2 preceding) as pressuremean,
        avg(vibration) over(partition by machineID order by machineID, datetime rows 2 preceding) as vibrationmean,
        row_number() over (partition by machineID order by machineID, datetime) as rn,
        machineID, datetime
from telemetry) rt
where rt.rn % 3 = 0 and rt.voltmean is not null
order by rt.machineID, rt.datetime

For more details please review the SQL code.

Based on my experience with predictive maintenance use cases, I have noticed that SQL rolling feature engineering was best suited for time series ordered data split by machine. For on-prem scenarios, now with SQL Server R Services, it also enables R enthusiasts to run their R code to do other data wrangling, model building and even scoring code from right within SQL Server. Overall, this ends up being more efficient as there is no data movement, and the computation ends up being scalable.

However, there are many other ways of operationalizing this type of feature engineering at scale. For example, R Server on HDInsight combines the functionality of R with the power of Hadoop and Spark, and Azure Data Lake Analytics now supports running R on petabytes of data. The power of can be put towards transforming raw sensor data into meaningful data that can be leveraged for machine learning applications to provide value back to the business.
Quelle: Azure