/* ---- Google Analytics Code Below */

Wednesday, August 18, 2021

Qualcomm Launches 5G and AI to Target Commercial Drones

Qualcomm 5G Drones using  AI for Commercial Connectivity Platform

Smartphones in the sky: Qualcomm launches first 5G and AI platform targeting commercial drones

After supplying computer chips for NASA’s Ingenuity helicopter on Mars, Qualcomm has unveiled next generation technologies to help power safe, beyond line-of-sight drone flights on Earth.

BY MIKE FREEMAN  AUG. 17, 2021 6:08 PM PT

Fresh off being in the limelight for its processors powering NASA’s Ingenuity helicopter on Mars, Qualcomm has rolled out its next-generation technologies for drones on Earth — aiming to drive adoption beyond hobbyists and into commercial industries.

The San Diego wireless technology company announced its Flight RB5 5G Platform on Tuesday. It is the first drone system to include 5G connectivity and artificial intelligence technology, which eventually could help enable autonomous, beyond-line-of-sight flights for such things as crop inspection, search and rescue, powerline/wind turbine monitoring, package delivery, mapping and so on.

In addition to the processing/connectivity platform, Qualcomm also launched its first-ever reference design kit for drone developers — essentially an aircraft body, rotors and other necessities to help companies develop drones faster and easier. ... '   

Combat Drones Catching Rides

 Saw this demonstrated in simulated concept, expect to see more swarm and group applications.

DARPA's New Combat Drones Could Catch a Ride from Other Aircraft   By Popular Science  February 23, 2021

Imagine an unmanned aircraft that is able to launch its own air-to-air weapon. That agile machine would itself first deploy from a bigger, crewed airplane, meaning that the entire system would involve missiles inside a drone that detaches from an airplane—like airborne Russian nesting dolls. This is the idea behind the aptly-named LongShot project, which the Defense Advanced Research Projects Agency, or DARPA, announced last week.

Details about the project are scant, but here's what we know about it.

DARPA said in a news release that the purpose was to create a new craft "that can significantly extend engagement ranges, increase mission effectiveness, and reduce the risk to manned aircraft."

In other words, the system is a type of "standoff" weapon—just like a B-52 bomber deploying a cruise missile while flying outside an adversary's airspace, it's an approach that keeps an aircraft and its crew far away from enemy air defenses.... 

The U.S. Defense Advanced Research Projects Agency last week announced the LongShot project to develop an unmanned aircraft able to launch its own air-to-air weapon ....

Credit: Senior Airman Haley Stevens /U.S. Air Force

From Popular Science

Saving the Grape Crop with Robots and AI

AI linking with sensors for agricultural analysis.

Robots and AI Help Save Multibillion Dollar Grape Crop   By Cornell University, August 18, 2021

 Researchers at Cornell University are using robotics and AI technology to identify grape plants infected with a damaging fungus that attacks wine grapes and other plants.

Adjunct professor Lance Cadle-Davidson developed prototypes of imaging robots as part of a team at the U.S. Department of Agriculture's Agricultural Research Service. The robots could scan grape leaf samples automatically, but researchers were bottlenecked by the need to manually assess thousands of grape leaf samples for evidence of infection.

Assistant research professor Yu Jiang and his team used AI to address the issue, applying deep neural networks to analyze microscopic images of grape leaves.

"It has revolutionized our science," Cadle-Davidson says. "Yu's AI tools actually do a better job of explaining the genetics of these grapes than we can do sitting at a microscope for months at a time doing backbreaking work."

The team describes its work in "Deep Learning-Based Saliency Maps for the Quantification of Grape Powdery Mildew at the Microscopic Level," which won a best paper award at the 2021 American Society of Agricultural and Biological Engineers annual international meeting.

From Cornell University

View Full Article

Updating Google Maps

 Impressive capabilities.

Three new ways anyone can update Google Maps

With all the change our world saw over the past year, people are relying more than ever on high-quality, updated information about the places around them — like if a nearby restaurant is open or if a local grocery store has updated hours.

To make sure your map reflects the real world, we make it easy for anyone with a Google account to contribute their local knowledge about more than 200 million places in Google Maps. These community-led updates help people everywhere make better decisions about what to do and where to go. And at the heart of the fresh information you see on Google Maps are Local Guides. This community of contributors who help others by updating Google Maps has reached a new milestone: 150 million Local Guides around the world. 

Today, we’re introducing three new Google Maps features that will help make sharing and finding local recommendations and information easier.  ... " 

Building TimeSeries Databases

Spent much of our careers working with timeseries data in supply chain analysis...

Not enough metadata considered?    

Always-on Time-Series Database: Keeping Up Where There's No Way to Catch Up  By Theo Schlossnagle, Justin Sheehy, Chris McCubbin

Communications of the ACM, July 2021, Vol. 64 No. 7, Pages 50-56  10.1145/3442518

In all likelihood, you have never given so much as a thought to what it might take to produce your own database. And you will probably never find yourself in a situation where you need to do anything of the sort.

But, if only as a thought exercise, consider this for a moment: What if, as a core business requirement, you found you needed to provide for the capture of data from disconnected operations, such that updates might be made by different parties at the same time—or in overlapping time—without conflicts? And what if your service called for you to receive massive volumes of data almost continuously throughout the day, such that you couldn't really afford to interrupt data ingest at any point for fear of finding yourself so far behind present state that there would be almost no way to catch up? Given all that, are there any commercially available databases out there you could use to meet those requirements?

Right. So, where would that leave you? And what would you do then? We wanted to explore these questions with Theo Schlossnagle, who did, in fact, build his own time-series database. As the founder and CTO of Circonus, an organization that performs telemetry analysis on an already large and exponentially growing number of IoT (Internet of Things) devices, Schlossnagle had good reason to make that investment.

Justin Sheehy, the chief architect of global performance and operations for Akamai, asks Schlossnagle about the thinking behind that effort and some of the key decisions made in the course of building the database, as well as what has been learned along the way. On behalf of ACM, Chris McCubbin, a senior applied scientist with Amazon Web Services, contributes to the discussion.

JUSTIN SHEEHY: As someone who once made the dubious decision to write my own database, I know it can prove to be the right thing to do, but—for most companies—I don't think it turns out that way. This isn't just a business question, but one that also has some interesting engineering dimensions to it. So, Theo, why did you feel the need to write your own time-series database?  .... 

Tuesday, August 17, 2021

Neuromarketing

 We touched on this years ago, as part of a broader cognitive science analysis of marketing,   it seems to have slipped behind the curtain since then.

A relook at the definition, there were a half dozen companies companies pushing it at the time. We tested some scenarios, mostly to test ad results.  It was common at the time to analyze Superbowl ads in context.

What is Neuromarketing?

It's time for an updated definition of "neuromarketing."  By Roger Dooley

Neuromarketing is the application of neuroscience and cognitive science to marketing.

Neuromarketing is the application of neuroscience and cognitive science to marketing. This can include market research that tries to discover customer needs, motivations, and preferences that traditional methods like surveys and focus groups can’t reveal. 

Neuromarketing can include the evaluation of specific advertising, marketing, packaging, content etc. to more accurately understand how customers react at the non-conscious level. And, it can include applying the knowledge obtained from neuroscience and cognitive science research to make marketing more effective without testing specific ads or other materials.

“Consumer neuroscience” is sometimes used as a synonym for neuromarketing.  ..."

Programmable Virus Traps?

Here on the face of it quite a claim.  Human tests still years away. 

Programmable Trap Can Kill Viruses  By R. Colin Johnson  in CACM

Commissioned by CACM Staff, August 17, 2021

Using computational genetic engineering, researchers at the Technical University of Munich (TUM) say they have invented a method of killing any type of virus. The researchers say they have demonstrated their solution on previously incurable hepatitis-B viruses, and are next aiming at the coronavirus.

The deoxyribonucleic acid (DNA) origami base-pair key-in-lock method they devised yields sphere-like icosahedral shells that kill viruses by clamping around each virion (the complete, infective form of a virus outside a host cell) until it is dead, dead, dead.

"Computational methods are an integral part of our work, from the initial design phase to the data processing for the final cryogenic electron microscopy," explains Hendrik Dietz,   a professor of biomolecular nanotechnology at TUM. "In one way or another, all members of my team are using computer modeling."

Shawn Douglas, a professor of molecular engineering at the University of California at San Francisco, said the work by Dietz and his team "is a remarkable demonstration of the power and versatility of DNA origami hierarchical assembly by non-covalent shape complementarity, a method that was also pioneered by the Dietz Lab. These new icosahedral shells are beautiful and inspiring artifacts of molecular engineering. I am excited to see how the team and their collaborators can adapt this system for in vivo applications in the coming years." 

The method uses a combination of software engineering, computational chemistry, and complementary DNA base-pair key-in-lock scaffolding to create icosahedral "traps" that snap shut around a target virus, isolating it harmlessly from both infecting and duplicating, and resulting in its eventual death.

The genetic engineering involved in the process does not tamper with human DNA; instead, the researchers use the universal building blocks of DNA—synthetic versions of the nucleotides cytosine, guanine, adenine, and thymine. A "virus trap" composed of locked-in-place isometric triangles is constructed in a spherical shape. A specific combination of nucleotides is first modeled in simulation to be the correct size to handle the target virus, then constructed using synthetic DNA from scientific laboratory suppliers. A coating is then inserted into the shell's interior to trap the virus inside, and a cap is added to block its exit. In practice, the virus victim's blood is flooded with the traps, which safely capture the individual virions and destroy them.

Dietz' TUM lab used a battery of software to model, simulate, validate, construct, and test prototype traps, the first of which were designed for hepatitis-B. The virus traps were tested in vitro (in a test tube or elsewhere outside the body) and were observed to be capable of trapping a targeted virus through the use of single-particle cryo-electron microscopy, which harnesses computationally intensive machine learning algorithms, according to Dietz. Subsequent in vivo (within a living organism) testing on mice showed the DNA-origami traps were capable of targeting individual virions inside the body, disarming them without disrupting bodily functions, and finally destroying them with natural immunological mechanisms. Human trials, however, are years away.    ... '

Quantum Error Correction

Brought back to my attention, with lots of links  to more detail,   From Google Research, Technical.

Demonstrating the Fundamentals of Quantum Error Correction

Wednesday, August 11, 2021

Posted by Jimmy Chen, Quantum Research Scientist and Matt McEwen, Student Researcher, Google Quantum AI

The Google Quantum AI team has been building quantum processors made of superconducting quantum bits (qubits) that have achieved the first beyond-classical computation, as well as the largest quantum chemical simulations to date. However, current generation quantum processors still have high operational error rates — in the range of 10-3 per operation, compared to the 10-12 believed to be necessary for a variety of useful algorithms. Bridging this tremendous gap in error rates will require more than just making better qubits — quantum computers of the future will have to use quantum error correction (QEC).

The core idea of QEC is to make a logical qubit by distributing its quantum state across many physical data qubits. When a physical error occurs, one can detect it by repeatedly checking certain properties of the qubits, allowing it to be corrected, preventing any error from occurring on the logical qubit state. While logical errors may still occur if a series of physical qubits experience an error together, this error rate should exponentially decrease with the addition of more physical qubits (more physical qubits need to be involved to cause a logical error). This exponential scaling behavior relies on physical qubit errors being sufficiently rare and independent. In particular, it’s important to suppress correlated errors, where one physical error simultaneously affects many qubits at once or persists over many cycles of error correction. Such correlated errors produce more complex patterns of error detections that are more difficult to correct and more easily cause logical errors.  ... " 

Robot Apocalypse

Recalled some of the points made here.   Some of our plant machinery was dated, and not likely to be replaced soon.  Yet also included advanced 'robotics' and controllers for parts of the process.  Depends on the industry and how the automation fits.  

Robot Apocalypse Hard to Find in America's Small, Mid-Sized Factories

Reuters, Timothy Aeppel, August 2, 2021

Although analysts have warned that millions of blue-collar jobs in the U.S. industrial heartland will soon be displaced by robots, that is not yet the case at small and medium-sized factories. Massachusetts Institute of Technology (MIT) researchers studied 34 companies with 500 or fewer employees in Ohio, Massachusetts, and Arizona, and found just one had acquired a significant number of robots in the past five years. MIT's Anna Waldman-Brown said, "The bulk of the machines we saw were from before the 1990s," and many older machines were upgraded with new computer controllers. Other companies have purchased advanced equipment like computer-guided cutting machines and inspection systems, but not robots, the researchers found, because smaller companies lack the money for robots or the skilled workers necessary to operate them.... 

Robotic Surgeries Better?

Useful to see some measurement of quality

Are Robotic Surgeries Really Better?

By The New York Times, August 17, 2021

Surgical procedures performed with the aid of a robot is sometimes marketed as the "best" form of surgery. But a recent review of 50 randomized controlled trials, testing robot-assisted surgeries against conventional methods for abdominal or pelvic procedures, suggests that while there may be some benefits to robotic surgery, any advantages over other approaches are modest.

Robotic surgery is performed by surgeons, not robots. But instead of conventional hand-held tools used in laparoscopic surgery, which involves tiny incisions, and open surgery, in which the surgeon enters the body through a large incision, the doctor uses a machine. The surgeon controls the machine's tools remotely by using joysticks and foot controls while viewing the surgical site through a high-definition monitor that provides a three-dimensional image of the proedure.

Some surgeons believe that these robots allow more precision during the operation, shorter recovery time, and generally better clinical outcomes for patients. But the review found that in many ways, compared outcomes from the robotic and conventional procedures showed little difference.

Article

Monday, August 16, 2021

IBM Using Digital Twins to Optimize AI

Spent years using simulation to optimize business process,  the idea of a Digital Twin came along to better utilize sensor context to leverage better solutions.  IBM is pushing it towards AI ... reading now.

How IBM is using digital twins to optimize AI

George Lawton, @glawton    in VentureBeat

Everyone is familiar with IBM’s leadership in IT, AI, and cloud services. It also happens to be one of the leading providers for Enterprise Asset Management software through its Maximo line of software and services. These tools help manage large machines, like factories, powerplants, and heavy equipment.

Now, with the rise of digital twins, IBM is pivoting this business as an onramp to bringing intelligence, agility, and efficiency to a wide range of industries. The company has gone so far as to declare “You cannot have AI without Digital Twin.” The new IBM digital twin exchange promises to create an app store for the digitization of the physical world that brings together enterprises and various services and tools providers.  ... ' 

11 Examples of Blockchain Use

Lisa Morgan  in InformationWeek  ... 

Blockchain Gets Real Across Industries

Blockchain use cases are exploding across industries because these days trust is 'a thing.' Here are 11 examples.

Bitcoin arguably put blockchain technology on the map for the general public. Of course, without blockchain technology, Bitcoin wouldn't exist.

Few remember the early "digital cash" initiatives that emerged not long after the debut of the World Wide Web. Back then, the fatal flaw was operating outside the established financial ecosystems. Blockchain, because it provides permanent, immutable records of transactions, was apparently the missing piece.

While many organizations have launched blockchain initiatives, multinational professional services network KPMG is now approaching it differently.

"Blockchain is now at a point where it's just part of the overall technology stack and that's how we're approaching it," said Tegan Keele, managing director of enterprise innovation at KPMG.

Up until about a month ago, Keele's team was branded as "the blockchain team." Part of that team focuses on crypto asset services that are about making blockchain useful for large financial services companies and helping crypto-native companies understand regulatory control Know Your Customer (KYC) transaction monitoring. Keele's focus is climate monitoring -- specifically, climate accounting that enables companies to prove their emissions footprint and their progress toward Net Zero and other climate-related goals.  ... ' 

Are You too Small For AI?

Given the author, you can predict the answer. Partially agree. But saw the same things early on for 'simpler' classic applications like analytics, which have some of the same issues with complexity, decision understanding, data availability, etc.  Worth a read, some good insight on application aspects. 

AI Doesn’t Have to Be Too Complicated or Expensive for Your Business  by Andrew Ng in HBR

Summary.   

For most companies that are interested in using AI, there isn’t a clear model to follow. The approach to building AI used by massive internet companies like Amazon and Google just doesn’t translate — most companies don’t have overflowing troves of data they can use to train models. So, industries such as manufacturing, agriculture, and healthcare need to take a different approach: programming with data, not code. Companies in these industries typically have relatively small data sets, face high costs for customizing a system, and are scared off by long gaps between pilot and product. But, given advances in AI technology, these organizations should shift their focus from building the right model — a software-focused approach — to focusing getting good data, which clearly illustrates the concepts we need the AI to learn, and using new machine learning operations (MLOps) tools. These tools that are geared to help produce high-quality datasets, in particular, hold the key to addressing the challenges of small datasets, high cost of customization, and the long road to getting an AI project into production outlined above. Companies should focus on gathering high-quality data, shifting the focus of their engineering corps away from model-centric approaches, and make the deployment process and MLOps tools needed to support it a central part of the planning project for any AI  ...  

Despite the vast potential of artificial intelligence (AI), it hasn’t caught hold in most industries. Sure, it has transformed consumer internet companies such as Google, Baidu, and Amazon — all massive and  data-rich with hundreds of millions of users. But for projections that AI will create $13 trillion of value a year to come true, industries such as manufacturing, agriculture, and healthcare still need to find ways to make this technology work for them. Here’s the problem: The playbook that these consumer internet companies use to build their AI systems — where a single one-size-fits-all AI system can serve massive numbers of users — won’t work for these other industries.  ... ' 

Simulating Building Energy use in Context with Digital Twins

A good example of the use of 'digital twins'.    Challenge will be to get the required detail for so many buildings with different ages, designs, updates, designs. Shows both the possibilities and the complexity involved.

ORNL's Simulation Tool Creates Digital Twins of Buildings From Coast to Coast

Oak Ridge National Laboratory, August 11, 2021 

Modeling software designed by researchers at the U.S. Department of Energy's Oak Ridge National Laboratory (ORNL) can simulate the energy profiles of all 129 million buildings in the U.S. The researchers developed the Automatic Building Energy Modeling (AutoBEM) software suite to realize ORNL's five-year Model America vision of producing digital twins for all U.S. buildings. AutoBEM uses high-performance computing to collect publicly available data like satellite imagery and street views, then generates a building energy model to forecast which energy-saving technologies could be implemented. ORNL's Joshua New said, "Before this program, no one had the capability to perform that analysis with detailed, building-specific energy modeling at this scale. Individual utilities now have the capability to perform modeling to show the potential of reducing demand and greenhouse gas emissions."

Overview of AI Labs in India

 An description and overview of Indian AI Labs via ODSC

Standout Artificial Intelligence Labs in India

ODSC - Open Data Science

Our passion is bringing thousands of the best and brightest data scientists together under one roof for an incredible learning and networking experience.

As dynamic fields, what we can do with and how we apply data science and artificial intelligence is almost constantly changing and growing. Research institutions, in particular, contribute to the discovery of new innovations and applications. Right now, we are seeing a significant amount of interesting research coming out of institutions in India, some of which are discussed below in this list of artificial intelligence labs in India.  ... '

MUM a new way to search online.

Is MUM just like a conversation online?   Most interesting. I  just recently noticed the amount of searching I had to do just to get background info.  

This new AI tool from Google could change the way we search online  by 7wData

What does the future of internet search look like? Google envisions it as looking more like a casual conversation with a friend. 

While Google’s search engine has been online for over two decades, the technology that powers it has been constantly evolving. Recently, the company announced a new artificial-intelligence system called MUM, which stands for Multitask Unified Model. MUM is designed to pick up the subtleties and nuances of human language at a global scale, which could help users find information they search for more easily or allow them to ask more abstract questions.  

Google already used MUM in an independent task to learn more about the different ways people refer to COVID vaccines, but says that the new tech is not yet part of their search system. While there’s currently no set timeline on when the feature will roll out in live search, the team is actively working on developing other one-off tasks for MUM to complete. 

Here’s what to know about what MUM is, how it’s different from what’s come before, and more.   ... '

On the SAS and Microsoft Partnership

They write:

One year in: Six reasons you should pay attention to the SAS and Microsoft partnership  by NICK JOHNSON

Just over a year ago, SAS and Microsoft announced their strategic partnership. Since then we have been working together to provide the best experience and value to our customers as they migrate to the cloud. 

Here are six milestones you should know about; each highlights the early success of the two industry giants.  

The start of something big  

In February of 2021, SAS Viya – SAS’s cloud-native end-to-end analytics platform – launched on Microsoft Azure. This latest release of SAS is built to run seamlessly on SAS Cloud and Azure, making it easier to design, run, and manage analytics and data processing workloads in the cloud. 

A shared vision for a data-led future  

Satya Nadella CEO of Microsoft and Jim Goodnight CEO of SAS discussed the partnership in a webcast for SAS Global Forum 2021. The two discussed the potential for the SAS and Microsoft Partnership to shape the future of AI and analytics in the cloud.    ... ' 

Sunday, August 15, 2021

Exploratory Data Analyses

 Am a big proponent of exploratory data analysis, and making it initially as simple, visual and goal oriented as possible.   Here a piece from KDNuggets that shows one approach.  Link through.    Have read KDN for years, long before AI became big and we did what was called optimization and analytics. 

EDA is a fundamental early process for any Data Science investigation. Typical approaches for visualization and exploration are powerful, but can be cumbersome for getting to the heart of your data. Now, you can get to know your data much faster with only a few lines of code... and it might even be fun!   KNIME Fall Data Talks Bringing Business and Data Science Together

By Francois Bertrand, coder & designer for data visualization and games.

AI Learns by Doing More with Less

Interesting thoughts on the efficiency of natural and current AI systems. Biomimicry possibilities for future intelligences? 

AI Learns by Doing More with Less  By Washington University in St. Louis   from ACM   August 12, 2021

 A tiny insect brain has less than a million neurons but has a diversity of behaviors and is more energy efficient than current AI systems. These tiny brains serve as models for computing systems that are becoming more sophisticated as billions of silicon neurons can be implemented on hardware.   ... 

New research from Washington University in St. Louis finds that the ability of silicon neurons to learn to communicate and establish networks is key to producing artificial intelligence systems that are as energy-efficient as biological ones.

Researchers previously found that neurons in a computational system act as though they are embedded in a rubber sheet; the new study demonstrates how neurons learn to choose the most energy-efficient perturbations and wave patterns in the rubber sheet. Each neuron adjusts the electrical stiffness of the rubber sheet to vibrate the entire network in the most energy-efficient manner, using only local information.

A silicon neuron, researchers discovered, can test all communications routes at once and identify the most efficient way to connect to complete a specific task.  ... 

From Washington University in St. Louis  Full Article

Can a Robot Understand a Hug?

Connecting touch and sight sensory interactions. 

Robot Can Understand What a Hug Is  By New Scientist, February 9, 2021

A prototype soft robot with nylon skin developed by researchers at Cornell University is imbued with sensory perception that Cornell's Yuhan Hu said lies somewhere between human touch and sight.

The team stretched the skin over a 1.2-meter (3.9-foot) cylindrical scaffold atop a wheeled platform, with a commercial USB camera for interpreting different types of touch on the nylon.

The Cornell researchers then compiled a database from camera images of humans making one of six interactions with the robot's skin, training a neural network to detect and identify different interactions with up to 92% accuracy.

The team matched simple commands to the gestures, and also showed that with a projector added, the robot could display a user interface on its skin for use as a touchscreen...

Cornell University researchers have created a low-cost method for soft, deformable robots to detect a range of physical interactions, using a USB camera inside the robot to capture shadow movements of hand gestures on the robot's skin and classifying them with machine learning software... 

From New Scientist