Loading

Category: The Latest

1613 posts
On
Posted on
in category

Reliability concerns raised over pi-top’s STEM learning laptop

9 views
Comments are closed Reliability concerns raised over pi-top’s STEM learning laptop Comments are closed

TechCrunch has learned of a safety issue and a number of product reliability questions being raised about a modular computer made by a London edtech startup that’s intended for children to learn coding and electronics.

The product, called the pi-top 3, is a Raspberry Pi-powered laptop with a keyboard that slides out to access a rail for breadboarding electronics.

A student at a US school had to be attended by a nurse after touching a component in the device which had overheated, leaving them with redness to their finger.

A spokesperson for Cornell Tech confirmed the incident to us — which they said had happened in June. We’ve withheld the name of the school at their request.

In an internal pi-top email regarding this incident, which we’ve also reviewed, it describes the student being left with “a very nasty finger burn”.

Cornell Tech’s spokesperson told us it has stopped using the pi-top 3 — partly in response to this incident but also because of wider reliability issues with the device. They said some of their grad students will be working on a project with the K-12 team next semester with the aim of creating an alternative that’s more reliable, affordable and safe.

We have also been told of concerns about wider reliability issues with the pi-top 3 by a number of other sources.

We asked pi-top for comment on the safety incident at Cornell Tech and for details of how it responded. The company provided us with a statement in which it claims: “pitop incorporates all possible safeguards into our products to ensure they are safe.”

“As soon as we became aware of this incident we immediately investigated what had happened,” it went on. “We discovered that the incident was a one-in-a-million occurrence. The user dropped a piece of metal, with a specific size and shape, under the unit. This fell in such a way that it touched a particular pin and caused a linear regulator to heat up. They received a small minor burn to the tip of one finger when they tried to recover that piece of metal.”

“This is the only reported incident where a user has been hurt whilst using one of our products,” pi-top added.

It is not clear how many pi-top 3 laptops have been sold to schools at this stage because pi-top does not break out sales per product. Instead it provided us with a figure for the total number of devices sold since it was founded in 2014 — saying this amounts to “more than 200,000 devices in 4 years which have been used by more than half a million people”.

pi-top also says it has sold products to schools in 70 countries, saying “thousands” of schools have engaged with its products. (The bright green color of the laptop is easy to spot in promotional photos for school STEM programs and summer camps.)

The London-based DIY hardware startup began life around five years ago offering a ‘3D-print it yourself‘ laptop for makers via the Kickstarter crowdfunding platform before shifting its focus to the educational market — tapping into the momentum around STEM education that’s seen a plethora of ‘learn to code’ toys unboxed in recent years.

pi-top has raised more than $20M in VC funding to date and now sells a number of learning devices and plug-in components intended for schools to teach STEM — all of which build on the Raspberry Pi microprocessor.

pi-top adds its own layer of software to the Pi as well as hardware additions intended to expand the learning utility (such as a speaker for the pi-top 3 and an “inventors kit” with several electronics projects, including one that lets kids build and program a robot).

The pi-top 3 — its third device — was launched in October 2017, priced between $285-$320 per laptop (without or with a Raspberry Pi 3).

The distinctively bright green laptop is intended for use by students as young as eight years old.

Unusual failure mode

In the internal email discussing the “Cornell failure diagnosis” — which is dated July 16 — pi-top’s head of support and customer success, Preya Wylie, conveys the assessment of its VP of technology, Wil Bennett, that the “unusual failure mode was likely caused by an electrical short on the male 34-pin connector on the underside of the protoboard”.

She goes on to specify that the short would have been caused by the metal SD-card removal tool that’s bundled with the product — noting this was “reported to have been somewhere underneath the protoboard at the time”.

“[Bennett] has recreated the same conditions on his bench in China and has seen the pi-top enter similar failure modes, with an electrical short and subsequent overheating,” she writes.

An additional complication discussed in the email is that the component is designed to stay on at all times in order that the pi-top can respond to the power button being pressed when the unit is off. Wylie writes that this means, if shorted, the component remains “very hot” even when the pi-top has been shut down and unplugged — as heat is generated by the pi-top continuing to draw power from the battery.

Only once the battery has fully depleted will the component be able to cool down.

In the email — which was sent to pi-top’s founder and CEO Jesse Lozano and COO Paul Callaghan — she goes on to include a list of four “initial recommendations to ensure this does not happen again”, including that the company should inform teachers to remove the SD-card removal tool from all pi-top 3 laptops and to remove the SD card themselves rather than letting students do it; as well as advising teachers/users to turn the device off if they suspect something has got lost under the protoboard.

Another recommendation listed in the email is the possibility of creating a “simple plastic cover to go over the hub” to prevent the risk of users’ fingers coming into contact with hot components.

A final suggestion is a small modification to the board to cut off one of the pins to “greatly reduce the chance of this happening again”.

We asked pi-top to confirm what steps it has taken to mitigate the risk of pitop 3 components overheating and posing a safety risk via the same sort of shorting failure experienced by Cornell Tech — and to confirm whether it has informed existing users of the risk from this failure mode.

An internal pi-top sales document that we’ve also reviewed discusses a ‘back to school’ sales campaign — detailing a plan to use discounts to “dissolve as much pi-top [3] stock as we can over the next 8 weeks”.

This document says US schools will be targeted from mid August; UK schools/educators from early September; and International Schools Groups from early September. It also includes a strategy to go direct to US Private and Charter Schools — on account of “shorter decision making timelines and less seasonal budgets”.

It’s not clear if the document pre-dates the Cornell incident.

In response to our questions, pi-top told us it is now writing to pi-top 3 customers, suggesting it is acting on some of the initial recommendations set out in Wylie’s July 16 email after we raised concerns.

In a statement the company said: “Whilst it is highly unlikely that this would occur again, we are writing to customers to advise them to take a common-sense approach and switch off the unit if something has got lost inside it.  We are also advising customers to remove the SD card tool from the unit. These simple actions will make the remote possibility of a recurrence even less likely.”

In parallel, we have heard additional concerns about the wider reliability of the pi-top 3 product — in addition to the shorting incident experienced by Cornell.

One source, who identified themselves as a former pi-top employee, told us that a number of schools have experienced reliability issues with the device. One of the schools named, East Penn School District in the US, confirmed it had experienced problems with the model — telling us it had to return an entire order of 40 of the pi-top 3 laptops after experiencing “a large volume of issues”.

“We had initially purchased 40 pi-tops for middle level computers classes,” assistant superintendent Laura Witman told us. “I met one of the owners, Jesse, at a STEM conference. Conceptually the devices had promise, but functionally we experienced a large volume of issues. The company tried to remedy the situation and in the end refunded our monies. I would say it was learning experience for both our district and the company, but I appreciate how they handled things in the end.”

Witman did not recall any problems with pi-top 3 components overheating.

A US-based STEM summer camp provider that we also contacted to confirm whether it had experienced issues with the pi-top 3 — a device which features prominently in promotional materials for its program — declined to comment. A spokesman for iD Tech’s program told us he was not allowed to talk about the matter.

A separate source familiar with the pi-top 3 also told us the product has suffered from software reliability issues, including crashes and using a lot of processor power, as well as hardware problems related to its battery losing power quickly and/or not charging. This source, who was speaking on condition of anonymity, said they were not aware of any issues related to overheating.

Asked to respond to wider concerns about the pi-top 3’s reliability, pi-top sent us this statement:

pitop is a growing and dynamic company developing DIY computing tools which we believe can change the world for the better. In the past four and a half years we have shipped hundreds of thousands of products across our entire product range, and pitop hardware and software have become trusted assets to teachers and students in classrooms from America to Zimbabwe. pitop products are hard at work even in challenging environments such as the UN’s Kakuma refugee camp in Northern Kenya.

At the heart of our products is the idea that young makers can get inside our computers, learn how they work and build new and invaluable skills for the future. Part of what makes pitop special, and why kids who’ve never seen inside a computer before think it’s awesome, is that you have to build it yourself straight out of the box and then design, code and make electronic systems with it. We call this learning.

The nature of DIY computing and electronics means that, very occasionally, things can fail. If they do, pitop’s modular nature means they can be easily replaced. If customers encounter any issues with any of our products our excellent customer support team are always ready to help.

It is important to say that all electronic systems generate heat and Raspberry Pi is no exception. However, at pitop we do the very best to mitigate thanks to the cutting-edge design of our hardware. Faults on any of our products fall well below accepted thresholds. Although we are proud of this fact, this doesn’t make us complacent and we continually strive to do things better and provide our customers with world-class products that don’t compromise on safety.

Thousands of schools around the world recognise the fantastic benefits the pitop [3], pitop CEED, and pitop [1] brings as a Raspberry Pi-powered device. Our new flagship products, the pitop [4] and our learning platform, pitop Further, take coding education to the next level, as a programmable computing module for makers, creators and innovators everywhere. We are proud of our products and the enormous benefits they bring to schools, students and makers around the world.

Internal restructuring

We also recently broke the news that pi-top had laid off a number of staff after losing out on a large education contract. Our sources told us the company is restructuring to implement a new strategy. pi-top confirmed 12 job cuts at that stage. Our sources suggest more cuts are pending.

Some notable names departing pi-top’s payroll in recent weeks are its director of learning and research, William Rankin — formerly a director of learning at Apple — who writes on LinkedIn that he joined pi-top in March 2018 to “develop a constructionist learning framework to support pi-top’s maker computing platform”. Rankin left the business this month, per his LinkedIn profile.

pi-top’s chief education and product officer, Graham Brown-Martin — who joined the business in September 2017, with a remit to lead “learning, product design, brand development and communication strategy” to support growth of its “global education business, community and ecosystem” — also exited recently, leaving last month per his LinkedIn.

In another change this summer pi-top appointed a new executive chairman of its board: Stanley Buchesky, the founder of a US edtech seed fund who previously served in the Trump administration as an interim CFO for the US department for education under secretary of state, Betsy DeVos.

Buchesky’s fund, which is called The EdTech Fund, said it had made an investment in pi-top last month. The size of the investment has not been publicly disclosed.

Buchesky took over the chairman role from pi-top board member and investor Eric Wilkinson: A partner at its Series A investor, Hambro Perks. Wilkinson remains on the pi-top board but no longer as exec chairman.

The job cuts and restructuring could be intended to prepare pi-top for a trade sale to another STEM device maker, according to one of our sources.

Meanwhile pi-top’s latest device, the pi-top 4, represents something of a physical restructuring of its core edtech computing proposition which looks intended to expand the suggestive utility it offers teachers via multiple modular use-cases — from building drones and wheeled robots to enabling sensor-based IoT projects which could check science learning criteria, all powered by pi-top’s encased Raspberry Pi 4.

Out of the box, the pi-top 4 is a computer in a box, not a standalone laptop. (Though pi-top does plan to sell a range of accessories enabling it be plugged in to power a touchscreen tablet or a laptop, and more.)

pi top 4 4

pi-top is in the process of bringing the pi-top 4 to market after raising almost $200,000 on Kickstarter from more than 500 backers. Early backers have been told to expect it to ship in November.

While pi-top’s predecessor product is stuck with the compute power of the last-gen Raspberry Pi 3 (the pi-top 3 cannot be upgraded to the Raspberry Pi 4), the pi-top 4 will have the more powerful Pi 4 as its engine.

However the latter has encountered some heat management issues of its own.

The Raspberry Pi Foundation recently put out a firmware update that’s intended to reduce the microprocessor’s operating temperature after users had complained it ran hot.

Asked whether the Foundation has any advice on encasing the Raspberry Pi 4, in light of the heat issue, founder Eben Upton told us: “Putting the Pi in a case will tend to cause it to idle at a higher temperature than if it is left in the open. This means there’s less temperature ‘in reserve’, so the Pi will throttle more quickly during a period of sustained high-intensity operation.”

“In general, the advice is to choose a case which is appropriate to your use case, and to update firmware frequently to benefit from improvements to idle power consumption as they come through,” he added.

TechCrunch’s Steve O’Hear contributed to this report

Article Source

 

On
Posted on
in category

How Oculus squeezed sophisticated tracking into pipsqueak hardware

11 views
Comments are closed How Oculus squeezed sophisticated tracking into pipsqueak hardware Comments are closed

Making the VR experience simple and portable was the main goal of the Oculus Quest, and it definitely accomplishes that. But going from things in the room tracking your headset to your headset tracking things in the room was a complex process. I talked with Facebook CTO Mike Schroepfer (“Schrep”) about the journey from “outside-in” to “inside-out.”

When you move your head and hands around with a VR headset and controllers, some part of the system has to track exactly where those things are at all times. There are two ways this is generally attempted.

One approach is to have sensors in the room you’re in, watching the devices and their embedded LEDs closely — looking from the outside in. The other is to have the sensors on the headset itself, which watches for signals in the room — looking from the inside out.

Both have their merits, but if you want a system to be wireless, your best bet is inside-out, since you don’t have to wirelessly send signals between the headset and the computer doing the actual position tracking, which can add hated latency to the experience.

Facebook and Oculus set a goal a few years back to achieve not just inside-out tracking, but make it as good or better than the wired systems that run on high-end PCs. And it would have to run anywhere, not just in a set scene with boundaries set by beacons or something, and do so within seconds of putting it on. The result is the impressive Quest headset, which succeeded with flying colors at this task (though it’s not much of a leap in others).

What’s impressive about it isn’t just that it can track objects around it and translate that to an accurate 3D position of itself, but that it can do so in real time on a chip with a fraction of the power of an ordinary computer.

“I’m unaware of any system that’s anywhere near this level of performance,” said Schroepfer. “In the early days there were a lot of debates about whether it would even work or not.”

Our hope is that for the long run, for most consumer applications, it’s going to all be inside-out tracking.

The term for what the headset does is simultaneous localization and mapping, or SLAM. It basically means building a map of your environment in 3D while also figuring out where you are in that map. Naturally robots have been doing this for some time, but they generally use specialized hardware like lidar, and have a more powerful processor at their disposal. All the new headsets would have are ordinary cameras.

“In a warehouse, I can make sure my lighting is right, I can put fiducials on the wall, which are markers that can help reset things if I get errors — that’s like a dramatic simplification of the problem, you know?” Schroepfer pointed out. “I’m not asking you to put fiducials up on your walls. We don’t make you put QR codes or precisely positioned GPS coordinates around your house.

“It’s never seen your living room before, and it just has to work. And in a relatively constrained computing environment — we’ve got a mobile CPU in this thing. And most of that mobile CPU is going to the content, too. The robot isn’t playing Beat Saber at the same time it’s cruising though the warehouse.”

It’s a difficult problem in multiple dimensions, then, which is why the team has been working on it for years. Ultimately several factors came together. One was simply that mobile chips became powerful enough that something like this is even possible. But Facebook can’t really take credit for that.

More important was the ongoing work in computer vision that Facebook’s AI division has been doing under the eye of Yann Lecun and others there. Machine learning models frontload a lot of the processing necessary for computer vision problems, and the resulting inference engines are lighter weight, if not necessarily well understood. Putting efficient, edge-oriented machine learning to work inched this problem closer to having a possible solution.

Most of the labor, however, went into the complex interactions of the multiple systems that interact in real time to do the SLAM work.

“I wish I could tell you it’s just this really clever formula, but there’s lots of bits to get this to work,” Schroepfer said. “For example, you have an IMU on the system, an inertial measurement unit, and that runs at a very high frequency, maybe 1000 Hz, much higher than the rest of the system [i.e. the sensors, not the processor]. But it has a lot of error. And then we run the tracker and mapper on separate threads. And actually we multi-threaded the mapper, because it’s the most expensive part [i.e. computationally]. Multi-threaded programming is a pain to begin with, but you do it across these three, and then they share data in interesting ways to make it quick.”

Schroepfer caught himself here; “I’d have to spend like three hours to take you through all the grungy bits.”

Part of the process was also extensive testing, for which they used a commercial motion tracking rig as ground truth. They’d track a user playing with the headset and controllers, and using the OptiTrack setup measure the precise motions made.

Testing with the OptiTrack system.

To see how the algorithms and sensing system performed, they’d basically play back the data from that session to a simulated version of it: video of what the camera saw, data from the IMU, and any other relevant metrics. If the simulation was close to the ground truth they’d collected externally, good. If it wasn’t, the machine learning system would adjust its parameters and they’d run the simulation again. Over time the smaller, more efficient system drew closer and closer to producing the same tracking data the OptiTrack rig had recorded.

Ultimately it needed to be as good or better than the standard Rift headset. Years after the original, no one would buy a headset that was a step down in any way, no matter how much cheaper it was.

“It’s one thing to say, well my error rate compared to ground truth is whatever, but how does it actually manifest in terms of the whole experience?” said Schroepfer. “As we got towards the end of development, we actually had a couple passionate Beat Saber players on the team, and they would play on the Rift and on the Quest. And the goal was, the same person should be able to get the same high score or better. That was a good way to reset our micro-metrics and say, well this is what we actually need to achieve the end experience that people want.”

the computer vision team here, they’re pretty bullish on cameras with really powerful algorithms behind them being the solution to many problems.

It doesn’t hurt that it’s cheaper, too. Lidar is expensive enough that even auto manufacturers are careful how they implement it, and time-of-flight or structured-light approaches like Kinect also bring the cost up. Yet they massively simplify the problem, being 3D sensing tools to begin with.

“What we said was, can we get just as good without that? Because it will dramatically reduce the long term cost of this product,” he said. “When you’re talking to the computer vision team here, they’re pretty bullish on cameras with really powerful algorithms behind them being the solution to many problems. So our hope is that for the long run, for most consumer applications, it’s going to all be inside-out tracking.”

I pointed out that VR is not considered by all to be a healthy industry, and that technological solutions may not do much to solve a more multi-layered problem.

Schroepfer replied that there are basically three problems facing VR adoption: cost, friction, and content. Cost is self-explanatory, but it would be wrong to say it’s gotten a lot cheaper over the years. Playstation VR established a low-cost entry early on but “real” VR has remained expensive. Friction is how difficult it is to get from “open the box” to “play a game,” and historically has been a sticking point for VR. Oculus Quest addresses both these issues quite well, being at $400 and as our review noted very easy to just pick up and use. All that computer vision work wasn’t for nothing.

Content is still thin on the ground, though. There have been some hits, like Superhot and Beat Saber, but nothing to really draw crowds to the platform (if it can be called that).

“What we’re seeing is, as we get these headsets out, and in developers hands that people come up with all sorts of creative ideas. I think we’re in the early stages — these platforms take some time to marinate,” Schroepfer admitted. “I think everyone should be patient, it’s going to take a while. But this is the way we’re approaching it, we’re just going to keep plugging away, building better content, better experiences, better headsets as fast as we can.”

Article Source

 

On
Posted on
in category

Apple reportedly launching new iPhone Pro and iPads with better cameras, 16-inch MacBook Pro and new AirPods

15 views
Comments are closed Apple reportedly launching new iPhone Pro and iPads with better cameras, 16-inch MacBook Pro and new AirPods Comments are closed

Apple is getting ready for its usual fall iPhone launch event, which is rumored to be happening September 10, though the event hasn’t been officially confirmed this year. A new report from Bloomberg offers a preview of the lineup of hardware products it’s looking to debut this year. There are new iPhones, of course, including a new iPhone Pro model that replaces the XS line and adds a third, wider angle rear camera (which has been rumored previously), and a refreshed iPhone XR at the entry level that will also get a second, optical zoom camera.

These new iPhone Pros would pack a lot of other updates besides, though they’ll look visually similar beyond the changed camera module. They’ll offer wireless charging for AirPods with the Qi-enabled wireless charging case, for instance, for a quick top-up when you’re the road, and they’ll also get new matte finishes on some models vs. the glossy look common to all iPhone models today. Updated Face ID will offer unlocking at more angles, and they’ll pack “dramatically” better water resistance, as well as improved shatter resistance to shrive drops.

Also new this year, though not necessarily debuting at the same event, will be a new MacBook Pro with a display size somewhere over 16-inches, which Bloomberg reports will still manage to be similar overall in physical footprint to the current 15-inch MacBook Pros, thanks to a new bezel. There are also plans to roll out new AirPods, with a higher price tag but also added water resistance and noise cancelling features that the current AirPods lack.

On the iPad side, Apple will refresh its iPad Pro this year, with updated versions of the 11-inch and 12.9-inch models that will get spec bumps, plus better cameras, but otherwise remain the same in terms of form factor. The entry-level iPad will also get an update, with a screen size increase from 9.7-inches to 10.2-inches, which could mean that it also slims down its bezel and does away with the dedicated Home button, though the Bloomberg doesn’t make mention of how it will actually change to accommodate the larger display size.

Apple Watch will also be updated, with the same case design introduced last year, but with at least new case finishes, which have leaked via the watchOS 6 update as coming in titanium and ceramic.

Other planned updates in the report include details about the iPhone to follow in 2020, which it says will a rear-facing 3D camera, as well as 5G network support. The HomePod will also apparently get a sequel next year – a smaller version that will likely be a lot more affordable vs. the current $300 speaker.

Article Source