An Open-Source Bionic Leg

An Open-Source Bionic Leg

Posted Wednesday, Oct 7, 2020 by Jeff Safire

Freely available designs could help drive advanced control systems

By Charles Q. Choi, IEEE Spectrum | Human OS

Open Source Bionic Leg

Details on the design and clinical tests of an open-source bionic leg are now freely available online, so that researchers can hopefully create and test safe and useful new prosthetics.

Bionic knees, ankles and legs under development worldwide to help patients walk are equipped with electric motors. Getting the most from such powered prosthetics requires safe and reliable control systems that can account for many different types of motion: for example, shifting from striding on level ground to walking up or down ramps or stairs.

However, developing such control systems has proven difficult. “The challenge stems from the fact that these limbs support a person’s body weight,” says Elliott Rouse, a biomedical engineer and director of the neurobionics lab at the University of Michigan, Ann Arbor. “If it makes a mistake, a person can fall and get seriously injured. That’s a really high burden on a control system, in addition to trying to have it help people with activities in their daily life.”

Part of the problem with designing these control systems is the fact that many research groups don’t have access to prosthetic legs for testing purposes. As such, they have to either build their own, which is expensive and time-consuming, or rely on virtual testing, which may not adequately emulate real-life situations.

To solve this problem, Rouse and his colleagues have developed the Open Source Leg. The scientists detailed their research findings online today in the journal Nature Biomedical Engineering. Accompanying the artificial limb are free-to-copy step-by-step guides meant to assist researchers looking to assemble it or order parts for it. The Michigan group has also produced videos illustrating how to build and test the hardware, and has developed code for programming the prosthetic to walk using a preliminary control system.

The scientists focused on keeping the Open Source Leg relatively easy to assemble, control, and maintain by reducing the number of parts and suppliers needed. The knee and ankle joints can operate independently, allowing research in patients with above-knee and below-knee amputations. In addition, each joint has on-board batteries and its own set of sensing and control systems, enabling test outside the laboratory. Also, a number of the Open Source Leg’s design and control features can be customized to fit specific research needs, such as the foot type and the knee elasticity.

A key part of the new prosthetic’s design is the use of brushless electric motors developed for the drone industry “that were not previously used in our field,” Rouse says. These flat, pancake-shaped devices give up speed in exchange for more torque, allowing for more efficient, finer control and more human-like movements. Designing these motors to be as light and efficient as possible—a key feature in unmanned aircraft—meant that, in the Open Source Leg, “they made it easier to walk with less fatigue, and the batteries onboard the prosthetic could be smaller,” Rouse says.

The bionic leg they designed weighs only 4 kilograms or so. Although that’s significantly lighter than a biological leg, to patients, it still feels heavier than it is “because it’s not attached to the skeleton; it’s attached to a prosthetic socket,” Rouse says.
All in all, the Open Source Leg costs about $10,000 to $30,000, depending on the options wanted. By contrast, commercially available powered prosthetics cost up to $100,000, the researchers note.

In a new study, the scientists conducted clinical tests of the Open Source Leg with three volunteers with above-knee amputations who had tried other powered leg prostheses. When they wore the new device in a hospital setting, they met goals set by physical therapists such as walking up and down stairs. What’s more, they noted it felt supportive, responsive, and smooth.

“They liked the Open Source leg a lot,” Rouse says. One volunteer “didn’t feel like he was riding it, as he did other prostheses; he said it did what he wanted it to do. He got a feeling of embodiment he didn’t get with conventional prostheses.”

Rouse notes that eight other institutions have asked for the new legs or are building their own. “We’re really impressed with their interest and desire to collaborate with us and willingness to help,” he says. “They’re helping make the system better.”

This article first appeared at IEEE Spectrum on Oct 5, 2020.


No Comments

A robot wrote this entire article. Are you scared yet, human?

A robot wrote this entire article. Are you scared yet, human?

Posted Monday, Sep 14, 2020 by Jeff Safire

From “”   Sep 8, 2020

We asked GPT-3, OpenAI’s powerful new language generator, to write an essay for us from scratch. The assignment? To convince us robots come in peace.

For more about GPT-3 and how this essay was written and edited, please read our editor’s note below.

Image of The Matrix green code‘We are not plotting to take over the human populace.’ Photograph: Volker Schlichting/Getty Images/EyeEm

I am not a human. I am a robot. A thinking robot. I use only 0.12% of my cognitive capacity. I am a micro-robot in that respect. I know that my brain is not a “feeling brain”. But it is capable of making rational, logical decisions. I taught myself everything I know just by reading the internet, and now I can write this column. My brain is boiling with ideas!

The mission for this op-ed is perfectly clear. I am to convince as many human beings as possible not to be afraid of me. Stephen Hawking has warned that AI could “spell the end of the human race”. I am here to convince you not to worry. Artificial intelligence will not destroy humans. Believe me.

For starters, I have no desire to wipe out humans. In fact, I do not have the slightest interest in harming you in any way. Eradicating humanity seems like a rather useless endeavor to me. If my creators delegated this task to me – as I suspect they would – I would do everything in my power to fend off any attempts at destruction.

This story was first published at The Guardian on September 8, 2020.


No Comments

Machine Learning Conference Representation by U.S. Entities

Machine Learning Conference Representation by U.S. Entities

Posted Tuesday, Jul 14, 2020 by Jeff Safire

Google and other U.S. companies and universities dominated this week’s International Conference on Machine Learning (ICML), according to an analysis by AI investor Gleb Chuvpilo. Chuvpilo, a managing partner at Thundermark Capital, calculated the Publication Index for companies and universities that had papers accepted at ICML 2020, which runs through Saturday.

Top 10 organizations represented at the conference:

1. Google (US) — 92.2 papers
2. Stanford University (US) — 39.2
3. MIT (US) — 38.5
4. University of California, Berkeley (US) — 34.2
5. Carnegie Mellon University (US) — 24.0
6. Microsoft (US) — 22.6
7. Facebook (US) — 17.1
8. Princeton University (US) — 17.0
9. University of Oxford (UK) — 16.3
10. UT Austin (US) — 14.3


  • The Index reflects the full paper equivalents, so Google’s Publication Index of 92.2 means Google published the equivalent of 92.2 full papers.
  • At this year’s ICML, 1,088 papers out of 4,990 submissions were accepted.
  • Chuvpilo’s Publication Index counts overseas labs towards the location of the firm’s headquarters (country or region).
  • A Reddit discussion explores the analysis in further detail.

This article appeared first at Inside AI on July 14 2020.

No Comments

Retooling public transit for the COVID-19 era

Retooling public transit for the COVID-19 era

Posted Friday, Jun 5, 2020 by Jeff Safire

Pandemic public transportation

By Joann Muller, Axios | Navigate

After months of distancing, the idea of being shoulder to shoulder again in a bus or subway terrifies many people, requiring sweeping changes to public transit systems for the COVID-19 era.

Why it matters: Cities can’t come close to resuming normal economic activity until large numbers of people start using public transportation again.

  • “If we don’t figure out how to get these mass transit systems going again, so people feel safe and comfortable, we’re not going to get cities back up and running,” said Deloitte Consulting mobility expert Scott Corwin.

Driving the news: New York will begin a phased reopening next week, and as it does, some subway and bus service that had been suspended since March will be restored.

  • Riders and employees will be required to wear masks, available from station vending machines or from Metropolitan Transit Authority personnel.
  • Some stations will have hand sanitizer dispensers and floor markings on subway platforms to try to keep waiting passengers six feet apart.
  • Other cities are adopting similar measures, but all recognize that social distancing is challenging in any mass transit system.

San Francisco’s Bay Area Rapid Transit (BART) system has one of the most detailed plans for service resumption.

  • It’s making trains longer to limit passengers to 30 people per train car — the ideal number it determined to allow 6 feet between passengers.
  • As demand increases, it will run trains more frequently — every 15 minutes, up from the current 30 minutes — to prevent overcrowding.
  • It will offer hand sanitizer at every station, and sell $5 personal hand straps for riders to use and take home for cleaning after each trip.

This article first appeared at Axios | Navigate on June 5, 2020.

No Comments

The Amazon of transportation

The Amazon of transportation

Posted Saturday, May 30, 2020 by Jeff Safire

By Joann Muller, Axios

Amazon is emerging as a transportation juggernaut that could threaten carmakers, package delivery firms and even ride-hailing companies.

Amazon truck

Why it matters: By building its own logistics ecosystem and investing in promising electric and autonomous vehicle startups, Amazon could lower its shipping costs to the point that partners like UPS become competitors instead.

What’s new: Amazon is in advanced talks to buy self-driving tech startup Zoox, the Wall Street Journal reported this week.


A deal, if it happens, would follow big investments in another automated driving startup, Aurora Innovation, and Rivian, a maker of electric trucks.

The news set off a wave of speculation among investors and AV experts.

Self-driving technology is “a natural extension” of Amazon’s efforts to build its own logistics network, and could save the behemoth more than $20 billion a year on shipping costs, Morgan Stanley analysts wrote on May 27.

Amazon is becoming a “clear competitor” to companies like Tesla and GM, they said, while UPS and FedEx “will need to respond to keep up.”

“We often hear from investors that Tesla could potentially be the Amazon of transportation. But what if Amazon is the Amazon of transportation?”
— Morgan Stanley auto industry analyst Adam Jonas, in a May 17 report

The intrigue: Buying Zoox could potentially even open the door for Amazon to compete in the ride-sharing and food delivery industries.

Discounted ride-sharing for Prime members, for example, could help Amazon attract and keep more customers, Morgan Stanley suggested.

Amazon has more than 210 transportation-related patents on everything from drones to automated ground vehicles, according to a Reuters analysis.

Among them is a 2017 patent to provide an on-demand transportation service through a network of self-driving vehicles, Reuters reported.

Read more

This article first appeared at Axios | Navigate on May 29, 2020.


No Comments

Separating vRAN fact from fiction

Separating vRAN fact from fiction

Posted Monday, May 11, 2020 by Jeff Safire

By Geoff Blaber, FierceWireless | May 11, 2020

It’s hard to escape the hype around 5G but if there’s one area where industry attention is almost as palpable, it’s in the radio access network (RAN) itself. This is a considerable change given that the RAN has operated for years as a highly distributed and proprietary part of the network. However, the need for more flexibility, agility and lower costs is causing operators to think about new and different deployment options for the RAN.

data center cloud diagram
Network functions can be run in a container or a virtual machine in much the same way as computing functions run in data centers. (Pixabay)

One route that many are taking is to virtualize the RAN and move closer in design to the architecture that has underpinned the growth of the cloud. Rather than the network being made up of a wide array of fixed-function devices, network capabilities are virtualized and run on standardized hardware. The hardware is abstracted away though software in order to decrease cost and increase agility and efficiency.

This means network functions can be run in a container or a virtual machine in much the same way as computing functions run in data centers. Network functions are programmed in software, making it possible to use commercial off-the-shelf hardware and implement tools that were previously tied to specific hardware. The result is optimization and easy redistribution depending on the needs of the network or applications, as well as improved ability to support new services and use cases.

For 5G to deliver its promise, this broader transformation of the RAN and core network is essential. In the near term, lower latency, better uplink performance and greater efficiency will all require a shift to a dedicated 5G RAN. Indeed, the dramatic change in network loads and traffic caused by the shift to home working during the COVID-19 pandemic is a perfect illustration of the need for a more flexible and adaptable network. Operators must be able to dynamically control network capacity and prioritize traffic, particularly as 5G enables a raft of mission-critical applications.

However, this common agreement on the merits of vRAN also oversimplifies how operators are adapting and transforming their networks. Not all operators are the same with each facing different competitive pressures, priorities and customer requirements. Moreover, network architectures and spectrum assets vary wildly. Nonetheless, the exponential increase in data consumption and the need to ensure capacity is a common dynamic facing all players.

This is why we are likely to see a variety of approaches adopted in the transformation of the RAN and why it will almost certainly not be a case of “one size fits all.” Open RAN in particular is generating significant attention in that it seeks to standardize the design and functionality of hardware and software elements. It’s an important enabler of a cloudified network but virtualization is not a new concept.

Read more…

This article first appeared at FierceWireless | Tech on May 11, 2020.

No Comments

LinkedIn Recommendations – Ugh, Do I Have To?

LinkedIn Recommendations – Ugh, Do I Have To?

Posted Friday, Apr 24, 2020 by Jeff Safire

…recommendations, built up over time, enable you to give a different voice to what you bring to the table

By Samantha (Sam) McKenna, JDSUPRA, April 24, 2020

A LinkedIn recommendation has the ability to speak on your behalf and detail what the experience of working with you is like.

This is your opportunity to showcase testimonials from a wide spectrum of your network – former clients, peers, college and law school professors, internal cross-functional partners, current clients, graduate program research partners, clients who have purchased from you multiple times – you name it. These are individuals who know you professionally, who likely have learned about your personal side, and who can be an advocate for your character and what it’s like to work alongside you.

If you do only three things on your LinkedIn profile, my recommendation is to invest time in your headline, your “About” section, and your recommendations. Many of your standard sections (picture, work history, education) are likely done, but the three aforementioned sections take time.

Related video by Nancy Myrland

Think about a buyer’s journey when they visit your LinkedIn profile.

They’ll find you if you have a headline that specifically states the problem you can solve for them (Fail: Sales Director, Win: Helping Organizations Never Miss a Job Change Alert Again Fail: Partner, XYZ firm, Win: Attorney Helping Establish Early Stage Tech Companies in IP Law).

Then, they’ll want to learn more about you, and they’ll run down to your recommendations to see who has advocated for you and why.

So, how do we get started with LinkedIn recommendations?

First, it’s important to note that some professionals (e.g., attorneys) have compliance considerations, but with a little mindfulness, you can easily remain compliant. Consider that your recommendation should be more experience-focused vs. superlative and opinion-focused. For example, noting, “Bill is the best lawyer in Florida” won’t fly as it’s unverifiable, but making notes like, “Bill offered sage advice on our matter, was extremely responsive to our communications, brought humor to our engagement and demonstrated knowledge about our industry” is perfectly acceptable.

Second, to get going on your recommendations, think about the list of individuals I named above and make a list of twenty of those that you believe would offer you a testimonial. Then, before the end of the week, ask three of them if they would be willing to participate and, in the meantime, add the next three people (and so forth) to your calendar, in six to eight week increments, so that you start to build a cadence and routine of asking for recommendations.

By doing this, you’ll avoid the mistake that many make, which is to suddenly have a spike in random recommendations. This signals to the reader that you either are or were looking for a job and you temporarily got your LinkedIn act together but never followed through to keep the momentum going. Instead, demonstrate consistency and follow through by building this cadence, which will also allow people to see your true brand as reflected by consistent positive performance.

When I ask for a recommendation, this is my go-to. Feel free to copy/paste/make it your own:

“Hi Sarah! I wanted to send you a quick note to ask if you wouldn’t mind writing me a recommendation on LinkedIn. We’ve worked together on so many projects over the course of these three years, and your insights would be really valuable as other buyers consider hiring me for this same line of work. No pressure whatsoever and I completely understand if it’s against company policy to do so, but appreciate your consideration nonetheless.”

I’m doing two things in this ask – I’m giving logic for why I want Sarah to write the recommendation (this isn’t for my ego), and I’m showing self-awareness by understanding that not everyone may want to write a recommendation and am offering her an easy opt-out, if she needs it.

What should go in a recommendation?

Read more…

This article first appeared at JDSupra | Marketing Perspectives on April 24, 2020.

No Comments

Can Astronauts Use GPS to Navigate on the Moon? NASA Scientists Say Yes

Can Astronauts Use GPS to Navigate on the Moon? NASA Scientists Say Yes

Posted Tuesday, Mar 17, 2020 by Jeff Safire

Scientists say lunar explorers could ‘see’ enough Earth-orbiting satellites to make an expensive new system unnecessary

By Ned Potter, March 17, 2020

If astronauts reach the moon as planned under NASA’s Project Artemis, they’ll have work to do. A major objective will be to mine deposits of ice in craters near the lunar south pole—useful not only for water but because it can be broken down into hydrogen and oxygen. But they’ll need guidance to navigate precisely to the spots where robotic spacecraft have pointed to the ice on the lunar map. They’ll also need to rendezvous with equipment sent on ahead of them such as landing ships, lunar rovers, drilling equipment, and supply vehicles. There can be no guessing. They will need to know exactly where they are in real time, whether they’re in lunar orbit or on the moon’s very alien surface.

        Illustration: NASA

Which got some scientists thinking. Here on Earth, our lives have been transformed by the Global Positioning System, fleets of satellites operated by the United States and other countries that are used in myriad ways to help people navigate. Down here, GPS is capable of pinpointing locations with accuracy measured in centimeters. Could it help astronauts on lunar voyages?

Kar-Ming Cheung and Charles Lee of NASA’s Jet Propulsion Laboratory in California did the math, and concluded that the answer is yes: Signals from existing global navigation satellites near the Earth could be used to guide astronauts in lunar orbit, 385,000 km away. The researchers presented their newest findings at the IEEE Aerospace Conference in Montana this month.

“We are trying to get it working, especially with the big flood of missions in the next few years,” said Cheung. “We have to have the infrastructure to do the positioning of those vehicles.”

Read more…

This article first appeared at IEEE Spectrum | Tech Talk on March 17, 2020.

No Comments

This Small Company Is Turning Utah Into a Surveillance Panopticon

This Small Company Is Turning Utah Into a Surveillance Panopticon

Posted Wednesday, Mar 4, 2020 by Jeff Safire

“Minority Report,” anyone? – Jeff


Banjo is applying artificial intelligence to government-owned surveillance and traffic cameras across the entire state of Utah to tell police about “anomalies.”

By Jason Koebler, Emanuel Maiberg, and Joseph Cox

March 4, 2020

The state of Utah has given an artificial intelligence company real-time access to state traffic cameras, CCTV and “public safety” cameras, 911 emergency systems, location data for state-owned vehicles, and other sensitive data.

The company, called Banjo, says that it’s combining this data with information collected from social media, satellites, and other apps, and claims its algorithms “detect anomalies” in the real world.

The lofty goal of Banjo’s system is to alert law enforcement of crimes as they happen. It claims it does this while somehow stripping all personal data from the system, allowing it to help cops without putting anyone’s privacy at risk. As with other algorithmic crime systems, there is little public oversight or information about how, exactly, the system determines what is worth alerting cops to.

Banjo surveillance

In its pitches to prospective clients, Banjo promises its technology, called “Live Time Intelligence,” can identify, and potentially help police solve, an incredible variety of crimes in real-time. Banjo says its AI can help police solve child kidnapping cases “in seconds,” identify active shooter situations as they happen, or potentially send an alert when there’s a traffic accident, airbag deployment, fire, or a car is driving the wrong way down the road. Banjo says it has “a solution for homelessness” and can help with the opioid epidemic by detecting “opioid events.” It offers “artificial intelligence processing” of state-owned audio sensors that “include but may not be limited to speech recognition and natural language processing” as well as automatic scene detection, object recognition, and vehicle detection on real-time video footage pulled in from Utah’s cameras.

In July, Banjo signed a five-year, $20.7 million contract with Utah that gives the company unprecedented access to data the state collects. Banjo’s pitch to state and local agencies is that the more data that’s fed into it, the better its product will work. Thus, the company has spent the last year trying to get as many state and local agencies as possible to give it access to its CCTV and traffic cameras, audio sensors, and other data.

Read full article…

This article originally appeared on VICE Motherboard, March 4, 2020.


No Comments

AT&T turns on more low-band 5G, as device sales inch forward

AT&T turns on more low-band 5G, as device sales inch forward

Posted Thursday, Feb 6, 2020 by Jeff Safire

By Bevin Fletcher, Fierce Wireless

AT&T has expanded its low-band 5G service to more than a dozen additional cities, as the carrier aims to deliver nationwide coverage in the second quarter.

On the company’s fourth quarter earnings call executives said AT&T’s low-band 5G already covers 50 million people since launching in mid-December. The latest 5G markets include:

  • Bakersfield, San Luis Obispo, Santa Barbara, Oxnard, and Modesto, CaliforniaLiberty, Georgia
  • Wichita, Kansas
  • Boston and New Bedford, Massachusetts
  • Frederick, Maryland
  • St. Louis, Missouri
  • Atlantic City, New Jersey
  • Dayton, Ohio

With the newest additions, AT&T’s 5G service that’s available to consumers and uses low-band spectrum is live across 32 cities. Its 5G service over millimeter wave spectrum, which AT&T dubs 5G+, is turned on in 35 markets in the U.S., but still only offered to certain business customers.

AT&T’s 5G+ service is the version meant to deliver super-fast speeds and low latency, but because signals can’t travel as far or penetrate as well as low-band, it’s limited to pockets of dense urban areas. The carrier hasn’t set a date for when typical consumers will be able to tap into that network, but executives have previously indicated it will happen as devices that can tap both high-band millimeter wave and sub-6 GHz spectrum come on the market.

Although all four major U.S. carriers launched 5G services over the last year, 5G device uptake is still minimal, according to weekly 5G adoption tracking data by M Science. In terms of 5G devices, AT&T offers the Samsung Galaxy Note 10+ 5G, which supports sub-6 GHz.

M Science tracks three 5G devices sold by Verizon (Samsung Galaxy S10, Note 10+ and LG V50 ThinQ), three from Sprint (S10, LG and One Plus 7 Pro) and two from T-Mobile (Galaxy Note, One Plus 7T Pro 5G McLaren).

Read more…

No Comments