árfolyamok

2020.12.04 19:00 UTC

BGN183,471Ft
CHF331,536Ft
CNY45,196Ft
CZK13,545Ft
DKK48,163Ft
EUR358,482Ft
GBP397,729Ft
HRK47,525Ft
JPY2,835Ft
NOK33,613Ft
PLN80,243Ft
RON73,598Ft
RSD3,049Ft
RUB3,996Ft
SEK34,979Ft
TRY37,783Ft
UAH10,434Ft
USD295,200Ft

hírek

Competition: Win an ADUM6421A eval board from ADI

2020-12-04 17:03 ElectronicsWeekly Alun Williams

There's a new competition on the site, give you the chance to win an ADI evaluation board. Its the ADUM6421A, a Quad-Channel Isolators with Integrated DC-to-DC Converter (3:1 directionality) evaluation board, to be precise.

This story continues at Competition: Win an ADUM6421A eval board from ADI

Or just read more coverage at Electronics Weekly

>>>


Dual-GbE router board offers PoE, 802.11ax, and an M.2 slot

2020-12-04 16:45 LinuxGizmos Eric Brown

Wally’s announced a compact “DR-6018-S” router board that runs Linux on a Qualcomm IPQ6010 with 802.11ax (Wi-Fi 6) and an M.2 slot plus dual GbE ports, including one with active PoE, and optional micro-SD, GPS, and USB 3.0. Wally’s Communications has followed up on its 5-port DR6018 v2 router board with a smaller DR-6018-S board […]

>>>


recycling 2d printers into 3d printing filamenet

2020-12-04 15:15 ElectronicsWeekly Steve Bush

Canon Ecology Industry has developed two types of 3d printing filament, made 100% from parts of discarded multi-function printer/scanners. These are the first in-house developed products from the company, which was set up in Japan to meet the sustainable development goals of Canon by recycling or arrange re-use of products made by the company. The ...

This story continues at recycling 2d printers into 3d printing filamenet

Or just read more coverage at Electronics Weekly

>>>


The Only Weapon We Have Is Innovation

2020-12-04 15:05 ElectronicsWeekly David Manners

2004 was another US election year and this had sensitivities for global companies like Xilinx, which had spread its R&D in centres around the world. ”In an election year it causes turmoil,” Xilincx CEO Wim Roelandts told me in February 2004, “people say we’re exporting jobs, or losing our edge at home.But you can’t innovate in ...

This story continues at The Only Weapon We Have Is Innovation

Or just read more coverage at Electronics Weekly

>>>


Most Read articles – Intel manufacturing, SMIC ban, China fabs

2020-12-04 12:36 ElectronicsWeekly Alun Williams

There's lunar developments, advanced semiconductor manufacturing in the US, a ban for SMIC, a distro deal and problems with memory fabs in China...

This story continues at Most Read articles – Intel manufacturing, SMIC ban, China fabs

Or just read more coverage at Electronics Weekly

>>>


Leonardo seeks increased diversity with AFBE-UK

2020-12-04 11:19 ElectronicsWeekly Steve Bush

Leonardo and the Association for Black and Minority Ethnic Engineers (AFBE-UK) have formed a partnership to increase diversity at the defence engineering company, which has over 7,500 employees across the UK, to “build an inclusive culture for all of its people and to help attract a richer diversity of job applicants to the business”. Dr Ollie ...

This story continues at Leonardo seeks increased diversity with AFBE-UK

Or just read more coverage at Electronics Weekly

>>>


PiFinger is a Fingerprint HAT for Raspberry Pi (Crowdfunding)

2020-12-04 10:55 CNXSoft Jean-Luc Aufranc (CNXSoft)

“There’s a HAT for that” they say, or something close to it… We’ve covered many...

>>>


Sundance joins AI Catapult incubator

2020-12-04 09:28 ElectronicsWeekly David Manners

Sundance Multiprocessor, the embedded module specialist, has joined Digital Catapult’s Machine Intelligence Garage business incubator, in a move that will accelerate and deepen its knowledge and expertise in the deployment of AI and MLtechniques. Sundance is part of an additional, new cohort of 10 early-stage and scaleup companies joining Digital Catapult’s Machine Intelligence Garage providing ...

This story continues at Sundance joins AI Catapult incubator

Or just read more coverage at Electronics Weekly

>>>


TTGO T5 4.7-inch e-Paper Display comes with ESP32 WiFi & Bluetooth SoC

2020-12-04 08:19 CNXSoft Jean-Luc Aufranc (CNXSoft)

We’ve very recently covered M5paper IoT development kit based on ESP32 WiSoC, and equipped with...

>>>


Strong rebound and double-digit growth for 2021

2020-12-04 07:30 ElectronicsWeekly David Manners

With estimated sales of $65.2 billion, DRAM was the biggest IC product category last year, says IC Insights, with NAND second biggest on $55.2 billion. DRAM has been the largest revenue-generating IC segment since 2017, even through the steep memory market downturn in 2019, and is forecast to retain is place as the largest IC ...

This story continues at Strong rebound and double-digit growth for 2021

Or just read more coverage at Electronics Weekly

>>>


Q3 equipment billings up 30% y-o-y

2020-12-04 07:08 ElectronicsWeekly David Manners

Semiconductor manufacturing equipment billings rose 30% y-o-y and 16% m-o-m to $19.4 billion, says SEMI in its Worldwide Semiconductor Equipment Market Statistics (WWSEMS) Report. Region Q32020 Q22020 Q32019 3Q2020 / 2Q2020 3Q2020 / 3Q2019 China 5.62 4.59 3.44 23% 63% Taiwan 4.75 3.51 3.90 36% 22% Korea 4.22 4.48 2.20 -6% 92% Japan 2.24 1.72 1.67 ...

This story continues at Q3 equipment billings up 30% y-o-y

Or just read more coverage at Electronics Weekly

>>>


700V and 800V SJ MOSFETs

2020-12-04 07:02 ElectronicsWeekly David Manners

MagnaChip Semiconductor of Korea has announced eight 700V and 800V series high-voltage Super Junction Metal Oxide Semiconductor Field Effect Transistors (SJ MOSFETs), optimised for TV, LED lighting and fast charger applications. MagnaChip began SJ MOSFET development in 2013. Cumulative shipments have now reached 1.5 billion units, and the company is now launching new generations of high-voltage ...

This story continues at 700V and 800V SJ MOSFETs

Or just read more coverage at Electronics Weekly

>>>


Mesa 20.3 released with Raspberry Pi 4 V3DK driver, Panfrost Bifrost support

2020-12-04 04:40 CNXSoft Jean-Luc Aufranc (CNXSoft)

We’ve previously reported that the Vulkan 1.0 conformant V3DK driver for Raspberry Pi 4 and...

>>>


Fishy

2020-12-04 02:50 ElectronicsWeekly David Manners

For some strange reason which has never been explained, the sub-£1bn fishing industry is said to be a key issue in the Brexit Trade Deal talks. Under the EU’s Common Fisheries Policy, all fleets have access to each other’s waters aside from the first 12 nautical miles away from the coast while EU ministers hold ...

This story continues at Fishy

Or just read more coverage at Electronics Weekly

>>>


Andes adds to its Linux-ready RISC-V line-up with L2 and multi-core ready models

2020-12-03 22:46 LinuxGizmos Eric Brown

Andes unveils four new Linux-focused RISC-V cores: The 32-bit A45MP and 64-bit AX45MP support up to 4x cores at up to 2.4GHz and offer optional L2 cache while the 32-bit A27L2 and AX27L2 also add optional L2. Last December when Andes Technology announced its RISC-V architecture AndesCore 27-series of Linux-ready CPU cores, we somehow missed […]

>>>


Smart cities: the case for lidar in intelligent transport systems

2020-12-03 21:25 Embedded.com Nitin Dahad

One aspect of the smart cities agenda is the deployment of intelligent transport systems. A pilot project using lidar sensors at the San Francisco Municipal Transportation Agency (SFMTA) has demonstrated how lidar has provided a solution for the city’s smart traffic signals pilot, a part of San Francisco’s Vision Zero policy.

The objective of the city’s Vision Zero policy is to improve road safety, since it is thought that every year about 30 people lose their lives and over 200 more are seriously injured while travelling on the streets of San Francisco. The objective of the smart traffic signals pilot within this larger framework was to explore the use of use of multimodal intelligent traffic signal systems (MMITSS), dedicated short range communication (DSRC), transit signal priority (TSP), and emergency vehicle preemption (EVP) technology to provide priority to emergency and transit vehicles. In addition, the MMITSS should also be able to detect pedestrians and bicycles to provide them leading intervals, scrambles, and/or protected phasing.

The first proof of concept which ended in January 2020 deployed lidar sensors at five intersections and demonstrated the ability to accurately and anonymously profile data, with 96% accuracy. The second proof of concept expects to add the data layer to the signal control network to enable the ‘intelligence’ in ITS; this is under way and expected to finish in early 2021.

We spoke two people involved who shed some light on the technology, its deployment and the results obtained, and here we present highlights of the conversations. First, we spoke with the lidar sensor technology provider Quanergy’s chief marketing officer, Enzo Signore. Then we dug deeper into the actual proof of concept project with Paul Hoekstra, who was the independent strategy execution consultant for the project with SFMTA.

Lidar: tracking IDs of objects anonymously

Enzo Signore of Quanergy explains the benefit of lidar technology in this type of application, requiring people and vehicle counting and flow management, and especially with the ban on facial recognition.

The key value proposition for lidar technology in applications like stadiums and smart cities is the ability to anonymously track an object travelling over various sensor points. For example, a car would go through many intersections, or a pedestrian will go through many areas. What Quanergy can do is to assign an ID to an object, and that ID will stay with the object throughout the journey in the area being monitored.

Quanergy basics of lidar
Lidar is a time-of-flight sensing technology that pulses low-power, eye-safe lasers and measures the time it takes for the laser to complete a round trip between the sensor and a target. The resulting aggregate data are used to generate a 3D point cloud image, providing both spatial location and depth information to identify, classify, and track moving objects. (Image: Quanergy)

This is very complex to do, because when you go through multiple intersections, you need multiple sensors and multiple servers for the edge compute. Most technologies would have a siloed view of only the area they are managing, and when crossing the boundary between one area and the other, the ID would be lost, and you’d be given another ID. With this approach, you start losing track of all the flow of people.

We have a technology called automated ID handover, that will pass the ID of the person or the vehicle, from one area to the other. So as long as we have field of view, the same ID will stay with the object. This gives very good end-to-end visibility and tracking. This can be important for airports, for example from kerb to gate, where you could optimize the passenger experience, and for shopping centers and cities. The single ID for each individual helps with enabling end-to-end analytics.

Quanergy’s sensors, the M- series, provide long range detection, such as the MQ-8, designed specifically for flow management applications. Here is how these sensors are different. Typical lidar sensors have a symmetric beam configuration. If you mount the sensor flat, then typically half the beam will go to the sky and half the beam will go to the ground. If it’s mounted 3 meters high on a street lamppost looking down for the pedestrian view, then in this configuration, half the beam is wasted.

In our design, all the beams are actually pointing down, which gives the ability to have a symmetric coverage of the ground. That means there are no blind spots when a person is walking across the field of view. This gives the ability to track without interruption a person or vehicle anywhere in the field of view. We can see an object up to a range of 70m (ie: 15,000 sq. m.). This is a very large area which would otherwise need many cameras to achieve a similar coverage. Hence this reduces the number of sensors and also the cost.

Overcoming privacy issues related to facial recognition

Paul Hoekstra, for SMTA, describes the thinking behind the implementation and the outcome of the first proof of concept (PoC) over five intersections on 3rd Street, and plans for expanding the coverage.

We started working with SFMTA, Cisco and Quanergy as the partners on this project in April 2019. Initially as part of the Cisco package we had DSRC sensors. We found we were using them just to listen to all the cars in the corridor and on the freeway that we were covering. We found that less than 1% of all cars actually broadcast that DSRC signal. From this use case perspective, the conclusion has to be that you cannot use DSRC for traffic flow measurement. It’s just not significant enough to make decisions on.

At this point in time, we’ve now completed the first (PoC) with Quanergy sensors, and now we are in the middle of the second PoC.

With the first PoC, we took 20 lidar sensors, installed them on five intersections on 3rd Street, near the new basketball stadium that opened last year. We had edge compute with Cisco TRX running the Quanergy QORTEX software. Data from the lidar goes via the TRX box and the Qortex software publishes the data to the network, which will go to the data center, a little VM cluster running Cisco Kinetic platform which stores all the messages – all 30 million a week.

Every Sunday reports are published on it, one for the vehicles, identifying the vehicle by the lidar ID at the intersection, with a whole bunch of attributes, like time, day of the week, where did it come from, where did it go, how often was there a stop, how long was the stop, what was the speed, was there an event (from the event calendar). In this way, we could connect all the intersections and follow vehicles through the corridor. And then we could say things like, “this is the amount that entered northbound at the southside of the corridor, and then how many turned off, and so on”.

With Quanergy’s QORTEX we have calibrated it and have achieved 96% accuracy. You can’t just count IDs with the lidar; you have to build logic to ensure that the same ID is in the egress as in the ingress. With that logic we can follow the cars through the intersection. We have logic that defines what a stop is. Hence that 96% accuracy is where we ended up, it is very precise. For pedestrians, with the zones that we defined, you can see whether the person is on or off the kerb. You can see if a person is inside or outside the boundaries of a crosswalk. You can see how close a car was to the person. With that kind of data, we can create reports of near misses. We’ve defined what near misses are – the vectors, the velocity, and then calculate the time for them to collide, and whether it is in a certain range, then you call it a near miss.

This is only possible because the data coming from QORTEX is so precise, that we can see without ever having to identify anyone. We are not storing any identifiable personal information. A person is just a dot. And a car is just a block, you don’t know what kind of car it is. We classify based on size.

SFMTA Intelliegent transport system pilot Paul Hoekstra
In the second proof of concept, 10 sensors cover a larger transport corridor in San Francisco, along 3rd Street between Channel and 20th Street (Image: Paul Hoekstra)

The first PoC was all about analysis. In the second one we are expanding the number of intersections. So now we have five, and we are going to go to 10. In other words, a larger corridor.

We’ll then layer in all the data that is currently locked inside the cabinets. In the cabinets there is a signal controller, and on the controller there are many actuators. These could be loop detecting a vehicle, a pedestrian pushbutton, or sensors in the light rail track. There is traffic signal prioritization. All of this sits in the embedded signal controller.

So what we are doing now is enabling two way information exchange with the signal controllers, taking all the data from the intersections, such as the lidar data and object classification (again completely anonymous), at the platforms and the bus stops. The object classification of those sensors (which is all processed on the sensors), will give us the number of people, as well as their classification – for example, is there somebody in a wheelchair, are there people pushing a stroller, or are they having a bike. Many of these factors will determine the dwell time of the transit vehicle. We want to know the predicted dwell time based on how many people are there.

From the back-end system we’re going to grab the number of vehicles. And then with the analytics we can determine whether we need say 20 seconds of dwell time or 32 seconds of dwell time. We can then extrapolate all the 10 intersections,

Treating the whole transport corridor as a network
To optimize the whole corridor, we can’t do it without knowing precisely where people and vehicles are and for how long. This means we treat the whole corridor as a network, not as an individual node. This means we run through the algorithms with high frequency, and now we are debating whether we need to go faster than 1 hertz, we recalculate everything every one second.

Then we actually tell the signal controller, you shall go green on northbound. That closes the loop. Learning from the supply chain work from Cisco, you know that that is the only way you can move stuff through the intersection. All the technology is available, but it only optimizes all the siloes. In this way, we’re taking a huge leap forward with a new paradigm, of integrated traffic management.

There are already 7,000 cameras in San Francisco. But cameras only give you a 2D picture. The precision of the location is less accurate than what you can achieve with lidar. Lidar always works, in rain, at night. And it stays very far away from the privacy issues. The moment that people know they are being tracked, or they can be recognized, there’s then the issue of people not having trust in the government to protect them.

The outcomes of this project are to enable emergency vehicles to have priority when dealing with emergencies, optimizing transit timings and stops, and even platooning cars up if no public transit available, to move them more effectively through the corridor.

Related Contents:

For more Embedded, subscribe to Embedded’s weekly email newsletter.

The post Smart cities: the case for lidar in intelligent transport systems appeared first on Embedded.com.

>>>


Working towards a sensor to monitor healing in chronic wounds

2020-12-03 18:41 ElectronicsWeekly Steve Bush

A flexible wearable sensor might one day guide treatment for chronic (long-lasting) wounds such as diabetic foot ulcers or pressure ulcers. Researchers from Russian institute Skoltech and the University of Texas at Austin have created a prototype for such a sensor, and demonstrated it in a simulate wound environment where it measured three biomarkers, all of ...

This story continues at Working towards a sensor to monitor healing in chronic wounds

Or just read more coverage at Electronics Weekly

>>>


Elektra Poll: From digital saxophones to Raspberry Pi 4, what’s the best example of Consumer Product Innovation?

2020-12-03 18:25 ElectronicsWeekly Alun Williams

All readers of Electronics Weekly are invited to take part in our poll and have your say on the destination of an Elektra 2020 Award. For the Consumer Product Innovation category, we are looking to readers to express their choice.

This story continues at Elektra Poll: From digital saxophones to Raspberry Pi 4, what’s the best example of Consumer Product Innovation?

Or just read more coverage at Electronics Weekly

>>>


Gate drivers have 3A, 6A and 9A output current

2020-12-03 17:42 ElectronicsWeekly David Manners

Infineon is shipping EiceDRIVER X3 Enhanced analog (1ED34xx) and digital (1ED38xx) gate driver ICs. These devices provide a typical output current of 3, 6 and 9A, precise short-circuit detection, a Miller clamp and soft turn-off. In addition, 1ED34xx offers an adjustable desaturation filter time and soft turn-off current with external resistors. These features combine to ...

This story continues at Gate drivers have 3A, 6A and 9A output current

Or just read more coverage at Electronics Weekly

>>>


AI is too powerful for engineers to handle alone

2020-12-03 17:25 ElectronicsWeekly Steve Bush

AI is too powerful for engineers to handle alone, is one of the conclusions of a study into autonomous systems and humanity people by the University of Texas at Austin. Instead, future designs must be guided by a broad range of societal stakeholders. Over 100 US autonomy experts were involved, whose work has been pulled together in ...

This story continues at AI is too powerful for engineers to handle alone

Or just read more coverage at Electronics Weekly

>>>


13MP Raspberry Pi 4 cam sells for $99

2020-12-03 17:14 LinuxGizmos Eric Brown

E-con announced two MIPI-CSI2 camera modules with Linux drivers: a 13MP, up to 4K “e-CAM130_CURB” for the Raspberry Pi 4 and an “e-CAM22_CUXVR” kit with an ultra-low-light, 2MP, HD camera for the Jetson AGX Xavier. E-con Systems has launched a pair of camera modules that connect to embedded Linux boards via MIPI-CSI2 interfaces. The $99 […]

>>>


Fable: The Acrobat Who Revolutionised Flight

2020-12-03 15:25 ElectronicsWeekly David Manners

An air force officer who was notorious for low flying and aerobatics was selected as one of the entrants in a competition to select a team to perform the “crazy flying” routine in the 1930 Royal Air Force Air Display at Hendon. He wrote off two aeroplanes during rehearsals. He went on to invent a ...

This story continues at Fable: The Acrobat Who Revolutionised Flight

Or just read more coverage at Electronics Weekly

>>>


Farnell opens Winter Games

2020-12-03 13:47 ElectronicsWeekly David Manners

Farnell has opened its Winter Games competition, giving customers access to exclusive weekly offers from market leading suppliers and fun online games with great prizes. Participants have the chance to win Amazon vouchers as well as the Winter Games grand prize, a £500 Apple voucher (or local currency equivalent). The Winter Games consists of two online ...

This story continues at Farnell opens Winter Games

Or just read more coverage at Electronics Weekly

>>>


Elektra 2020 – Vote for the University Research Readers’ Choice award

2020-12-03 13:39 ElectronicsWeekly Alun Williams

As we always say, this is your chance to shape the destination of another Elektra 2020 Award. We are looking for readers’ votes in the University Research category. In this category, visitors to the Electronics Weekly website are invited to select the University research project which they feel will make the largest impact on the ...

This story continues at Elektra 2020 – Vote for the University Research Readers’ Choice award

Or just read more coverage at Electronics Weekly

>>>


Graphene makes better heat pipes

2020-12-03 13:08 ElectronicsWeekly Steve Bush

Graphene can make heat pipes work much better, according to Chalmers University of Technology in Sweden. Heat pipes, often seen in high-end PCs, are sealed hollow tubes that contain a liquid at a pressure than ensures part of it is vapour at the temperatures of interest. If you put one end of a heat pipe ...

This story continues at Graphene makes better heat pipes

Or just read more coverage at Electronics Weekly

>>>


3d printing: less guilt when printing flexibles

2020-12-03 12:34 ElectronicsWeekly Steve Bush

How about a way to reduce guilt when 3d printing flexible objects – assuming you feel guilt at such times, of course. Creamelt is offering a potential route, by making flexible TP (thermoplastic polyurethane) filament out of old ski boots. Described as a sustainable alternative to virgin material, Creamelt TPU-R is 100% made from recycled ...

This story continues at 3d printing: less guilt when printing flexibles

Or just read more coverage at Electronics Weekly

>>>


Compact Jetson Xavier NX/ Nano open hardware baseboard supports Android

2020-12-03 11:42 CNXSoft Abhishek Jadhav

When it comes to NVIDIA Jetson family of modules, we should understand that NVIDIA Jetson...

>>>


PSU manufacture: Gresham buys Relec

2020-12-03 11:05 ElectronicsWeekly Steve Bush

US defence company Gresham Worldwide has bought Dorset PSU and display company Relec Electronics. It already owns Wiltshire PSU maker Gresham Power Electronics. “We spearheaded the Relec acquisition with a view to gaining access to additional resources and products to increase the range of offerings and service levels that Gresham Worldwide could provide to its customers,” ...

This story continues at PSU manufacture: Gresham buys Relec

Or just read more coverage at Electronics Weekly

>>>


HotDive converts your smartphone into a dive computer and underwater camera (crowdfunding)

2020-12-03 10:58 CNXSoft Jean-Luc Aufranc (CNXSoft)

Many recent phones are fitted with a pretty good camera, and some are waterproof, but...

>>>


Intel committed to advanced manufacturing in USA

2020-12-03 07:30 ElectronicsWeekly David Manners

Yesterday Intel CEO Bob Swan (pictured) committed the company to having advanced manufacturing in the USA. That would have gone without saying three years ago but, with TSMC’s rise to pre-eminence in process technology, there has been speculation that it might put leading edge product out to foundry. “We have a wonderful business model called ...

This story continues at Intel committed to advanced manufacturing in USA

Or just read more coverage at Electronics Weekly

>>>


Ayar demoes photonic Link on GloFo 45nm

2020-12-03 07:02 ElectronicsWeekly David Manners

Ayar Labs has demonstrated a monolithic electronic/photonic connection on Globalfoundries’ 45nm process. This is a step in providing chip-to-chip optical connectivity at scale for applications like AI, HPC, cloud, telecoms and aerospace. Over the last 18 months, Ayar Labs has been working with select semiconductor manufacturers, systems builders, and end users on co-design partnerships. The ...

This story continues at Ayar demoes photonic Link on GloFo 45nm

Or just read more coverage at Electronics Weekly

>>>


10kW from discrete GaN device

2020-12-03 07:01 ElectronicsWeekly David Manners

Transphorm is sampling its first 5th gen SuperGaNTM device meant for EVs, It claims to offer the world’s lowest packaged on-resistance with 25% lower power loss compared to SiC. ‘Transphorm’s demonstration of achieving 10 kilowatts of power from a discrete packaged GaN device in a bridge configuration is further validation of the exciting promise of ...

This story continues at 10kW from discrete GaN device

Or just read more coverage at Electronics Weekly

>>>


Qualcomm Snapdragon 888 specifications – Cortex-M1/A78/A55 Processor with Snapdragon X60 5G Modem

2020-12-03 06:03 CNXSoft Jean-Luc Aufranc (CNXSoft)

Qualcomm teased us with Snapdragon 888 5G mobile platform on the first day of the...

>>>


Vaccines!

2020-12-03 02:03 ElectronicsWeekly David Manners

Vaccines are here and who’s got ‘em? 9.8 billion doses had been bought by governments by November 20, says Duke University. Of that 9.8 billion, high-income countries have secured 3.49 billion doses, middle-income nations have obtained 828.8 million and lower middle-income nations will have access to 1.75 billion. Duke University researchers were unable to find ...

This story continues at Vaccines!

Or just read more coverage at Electronics Weekly

>>>


64-bit RISC-V core claims 10x better CoreMarks/Watt compared to other 3-5GHz CPUs

2020-12-02 19:48 LinuxGizmos Eric Brown

Micro Magic unveiled an up to 64-bit RISC-V core showing a groundbreaking 110,000 CoreMarks/Watt, with a 3GHz chip consuming less than 70mW. The company claims 10 times better CoreMarks/Watt compared to other processors in the 3-5GHz range. Considering the spectacular promise and sudden demise of AI tech firm Magic AI, we should perhaps be wary […]

>>>


3d printing: New handles for old

2020-12-02 18:51 ElectronicsWeekly Steve Bush

Another request from a friend – our wardrobe handles have fallen apart and are no longer available. Is black ok? Yes, so long as we can open the cupboard with it, and can we have four please. Luckily, although two of the originals had largely disintegrated (UV’s influence on ABS suspected), at least one was ...

This story continues at 3d printing: New handles for old

Or just read more coverage at Electronics Weekly

>>>


Off-line ac-dc converter stands by for

2020-12-02 18:17 ElectronicsWeekly Steve Bush

Viper31 is a mains voltage converter for ac-dc power supplies from STMicroelectronics which allows PSUs to be constructed that consume only 20mW no-load at 230Vac, while delivering up to 18W when configured as a universal (85 – 265Vac) fly-back converter. Also, consumption is <430mW at 230Vac with a 250mW load. It can be used in isolated ...

This story continues at Off-line ac-dc converter stands by for

Or just read more coverage at Electronics Weekly

>>>


Training AI models on the edge

2020-12-02 16:17 Embedded.com Haruyuki Tago

In recent years, a variety of companies have been introducing and promoting deep learning technology. However, two large issues are still to be solved when it comes to deep learning. One is the necessity to prepare a large amount of training data, and the other is the need for a huge amount of backpropagation and other calculations in the initial training stage. The latter was often performed on servers with high performing GPUs, located in the cloud, thus very power consuming. Therefore, it was not realistic to perform training on edge devices. However, a technology exists that can perform training and inference extracting features from only a small amount of data. This technology is known as Sparse Modeling.

In this article we will compare Sparse Modeling with deep learning; cover use cases where Sparse Modeling is beneficial; and explain the mechanism of Sparse Modeling training and inference. Furthermore, examples of visual inspection applications leveraging Sparse Modeling technology will be introduced. In this case, we will cover scratch detection on a flying crane pattern, which is one of the traditional patterns of Japanese paper.

What is Sparse Modeling

The word “sparse” is defined as “thinly dispersed or scattered”. Sparse Modeling is based on the assumption that essential information is actually very limited (therefore”sparsely distributed”). This technology identifies and extracts essential information from the input data for the output.

Sparse Modeling identifies the relationship between different data. When outputting, Sparse Modeling does not focus on the input data itself but focuses on the relationship between input and output data. By focusing on the relationship between the data, the quantity and quality of the input data itself does not matter. Consequently, only a small amount of data is needed. Sparse modeling is categorized as an unsupervised learning method in machine learning.

Deep learning generally delivers high performance in applications where sufficient data and annotations can be prepared (e.g., object detection for automated driving). However, Sparse Modeling expands the scope of AI application to use cases where large amounts of data cannot be collected, and interpretability is of high importance.

Comparison of Sparse Modeling and deep learning

Sparse modeling can handle both one-dimensional data such as time series and two-dimensional data such as images. Imaging applications include defect interpolation, defect inspection (anomaly detection), and super-resolution. Figure 1 shows a comparison between a conventional machine learning method and Sparse Modeling for a task of detecting defect images in solar panel inspections. You can see that the amount of training data (number of images) needed for Sparse Modeling is significantly smaller than that of deep learning. However, the accuracy of Sparse Modeling is more than 90% – with a training time of only 19 seconds.


Figure 1. Performance comparison in solar panel defect inspection by AI. (Source: Hacarus)

Applications where Sparse Modeling is beneficial

Sparse Modeling works even with small amounts of training data. This is because it extracts the features that are essential from the input data in the training stage. Looking at an example of a visual inspection application located in an industrial manufacturing site, there is often a large amount of “good” data, but very little bad data. This is a case where Sparse Modeling works as an AI solution.

Sparse modeling is computationally inexpensive for model creation, so it enables us not only to perform inference on edge devices, but also training.

When it comes to AI models performing both training and inference on edge devices, HACARUS calls this “True Edge”. By training on the edge, there is no need to send data to an external location (e.g., a server in the cloud) resulting in fewer data security concerns. The Congatec and HACARUS AI kit (Figure 2) is an example of a device capable of performing Sparse Modeling on the edge. This device is made from a Congatec Box-PC and a HACARUS AI Kit – which includes visual inspection software SPECTRO CORE.


Figure 2. Congatec Box-PC and Hacarus AI kit. (Source: congatec)

Training and Prediction with Sparse Modeling

The overall flow of visual inspection using Sparse Modeling is shown in Figure 3.


Figure 3. Overall flow of visual inspection system. (Source: Hacarus)

In this example, the objective is to detect a screw on the perforated metal as a foreign object. First of all, images of the perforated metal with no anomalies are prepared. In order to avoid errors due to differences in camera angles and target positions, it is recommended to prepare several dozen training images from slightly different viewpoints. In the training phase, these training images are input to the training algorithm to create an AI model. If the inspection target does not change, the AI model need only be created once.

In the prediction phase, we input the prediction image with a screw on it, as well as the AI model, into the inference algorithm to perform inference. As a result of the inference, an image with a red frame around the location of the foreign object is produced.

The details of the training phase are shown in Figure 4. First, the training image is divided into small segments called patches. The dictionary learning algorithm then analyzes the patches and extracts characteristic image patterns. In this example, 64 different patterns of holes in the perforated metal were extracted. Each of them is called a base, and the AI model (also called a trained dictionary) is a collection of 64 bases.

click for full size image

Figure 4. Training phase of anomaly detection. (Source: Hacarus)

The details of the prediction phase are shown in Figure 5. First, we broke down the prediction image (A) into patches of the same size as the training phase. For each decomposed patch, a base of the AI model is chosen and combined to find the combination that best represents the patch. This process is applied to all patches and stitched together to produce an approximate image using the AI model. We call it the reconstructed image (B).

click for full size image

Figure 5. Prediction phase of anomaly detection. (Source: Hacarus)

In the reconstructed image (B), the location of image features that are close to the training image, in this case a regular pattern of holes, is very similar to the prediction image (A). In contrast, locations with image features that are not included in the training image, i.e., locations where screws are placed, are not well represented by the basis included in the AI model. At that location, there is a difference between the reconstructed image and the prediction image. Thus, a foreign substance can be detected. The red area in the foreign object detection image (C) indicates the location of the foreign object.

Visual Inspection Applications Leveraging Sparse Modeling Technology

We conducted a scratch detection experiment using a flying crane image (image size 640 pixel x 480 pixel), a classic design for traditional Japanese paper. The experiment environment is shown in Figure 2.

Training

Thirteen normal images, shown in Figure 6, were used for training. This data set is less than a tenth of the images which would be required when using deep learning. Although the pattern of the cranes amongst the 13 images is the same, the position of the cranes in the images differ slightly. The reason for this is to provide a translation invariance to the AI model.


Figure 6. Training images of Japanese paper with flying crane pattern. (Source: Hacarus)

Example of Defect Detection

The prediction image is shown on the left side of Figure 7. Scratches are partially overlapped in the crane in the lower left corner. In the prediction image, the scratch is detected as a foreign object, surrounded by a red frame. Thus, the AI model was able to detect the change of the crane’s shape due to scratching, as it had already learned the shape of the crane.


Figure 7. One scratch defect on a crane. (Source: Hacarus)

Training Image Size and Training Time

Figure 8 is a graph showing training image size and training time for the flying crane pattern. In the case of 13 training images with a size of 1280 pixel x 960 pixel, the training time was only about 16 seconds. The training time is far shorter than that of deep learning. As the image size decreases, the learning time decreases almost in a linear form.

click for full size image

Figure 8. Training image size vs training time. (Source: Hacarus)

Prediction Image Size and Prediction Time

Figure 9 shows a graph of prediction image size and prediction time for the flying crane pattern. The training time is only about 0.6 seconds per 1280 pixel x 960 pixel prediction image size. As the image size decreases, the learning time decreases almost linearly to less than 0.1 seconds for 480 pixel x 360 pixel.

click for full size image

Figure 9. Prediction image size vs prediction time. (Source: Hacarus)

Detection Accuracy

Figure 10 shows the results evaluating a further 100 industrial images using SPECTRO, HACARUS’ visual inspection solution. We were able to obtain very positive results, with an accuracy of 96%, precision of 100% and a reproduction rate of 95.65%. The anomaly score in the graph is an internal SPECTRO indicator that takes a value from 0 to 1, and the higher the score, the higher the anomaly level. The horizontal axis shows the anomaly score and the vertical axis shows the cumulative percentage of the total score. The graph in red shows that the degree of abnormality is in the neighborhood of 0.133, which covers all defective products.


Figure 10. Accuracy results. (Source: Hacarus)

Conclusion

We covered the comparison of Sparse Modeling and deep learning, applications where Sparse Modeling is beneficial, and the mechanism of training and prediction in Sparse Modeling. Scratch detection of traditional Japanese paper patterns and flying cranes was successfully achieved by SPECTRO CORE, a visual inspection program that uses sparse modeling. The accuracy, precision, and reproducibility were confirmed to be high in this experiment.

Haruyuki Tago is Edge Evangelist at Hacarus, leading the Hacarus Edge team focused on building true Edge AI, capable of both training and inference on the edge. Mr. Tago brings over 40 years of international experience from senior management roles across firms such as Sony, Toshiba and IBM in microprocessor and SoC design management to Hacarus. He contributed greatly to the SoC designs which power the Sony PlayStations’ gaming consoles, as a member of the management team.

Related Contents:

For more Embedded, subscribe to Embedded’s weekly email newsletter.

The post Training AI models on the edge appeared first on Embedded.com.

>>>


Probe head supports wafer-level packages and known good die production

2020-12-02 16:12 ElectronicsWeekly Steve Bush

Smiths Interconnect has introduced a production probe head for wafer-level packages and known good die. Part of its Volta range Volta 180 is a WLCSP test connector for chips with 180um pitch IO. The series is capable of testing sorted die for engineering development or failure analysis, and is an alternative to cantilever and vertical probe ...

This story continues at Probe head supports wafer-level packages and known good die production

Or just read more coverage at Electronics Weekly

>>>


Xilinx Acquires Falcon Computing to Bring Adaptive Computing to More Developers

2020-12-02 15:50 Embedded.com Nitin Dahad

Xilinx’ acquisition of Falcon Computing Solutions announced this week will bring software programmability and adaptive computing to more software developers, by using Falcon Computing’s compiler technology to create custom high-performance accelerated applications using FPGAs and adaptive systems on chip (SoCs).

Falcon Computing Solutions is a provider of high-level synthesis (HLS) compiler optimization technology for hardware acceleration of software applications. Xilinx said the acquisition will make adaptive computing more accessible to software developers by enhancing the Vitis unified software platform with automated hardware-aware optimizations.

The integration of Falcon Computing’s compiler technology into the Vitis platform will allow software developers to accelerate C++ applications with minimal hardware expertise. Falcon Computing’s source code transformation reduces the need for application developers to adapt their code, or add architecture-specific programming directives, in order to achieve significant hardware acceleration.

“The growing demand for adaptive computing is driving a new era of FPGA adoption in the data center and embedded applications,” said Salil Raje, executive vice president and general manager for the data center group at Xilinx. “Falcon Computing’s innovative compiler technology and highly specialized compiler team will provide critical expertise that will advance software programmability and help bring the benefits of adaptive computing to more developers.”

The co-founder and chairman of Falcon Computing, Jason Cong, said, “Our compiler technology enables software developers to quickly achieve an order of magnitude acceleration over CPUs with very little knowledge of the FPGA hardware architecture, as our compiler provides a high degree of automation to optimize off-chip data movement, on-chip data reuse, memory partitioning, parallel and pipelined computation acceleration. The single-source Open-MP like programming style is very friendly to a large base of C/C++ software developers, especially those from the high-performance computing and embedded system communities.”

Falcon Computing merlin-diagram
To enable acceleration to software developers, the Merlin Compiler provides a familiar software flow with an automated C/C++ to FPGA. (Image: Falcon Computing Solutions)

Falcon Computing’s products include the Merlin Compiler, which addresses applications that involve parallel computing with a mix of specialized hardware such as multi-core CPUs, GPUs, and FPGAs. While hardware developers might have the expertise to deploy applications across these heterogeneous platforms, they are scarce as there is now a higher prevalence of software developers and data scientists in industry. To enable these software developers to perform the acceleration, the Merlin Compiler provides a familiar software flow with an automated C/C++ to FPGA. Hence Merlin bridges the gap between cognitive era applications and specialized hardware enabling software developers to enhance their application performance on heterogeneous platforms.

This is now the third company Cong has sold to Xilinx. As the Volgenau chair for engineering excellence at the Computer Science Department of UCLA, the director of the Center for Domain Specific Computing, a fellow of ACM and IEEE, and a member of the National Academy of Engineering, Cong co-founded Falcon Computing in 2014. With roots in academia and research, Falcon Computing has been at the forefront of the new wave of FPGA adoption. Previously, Cong also co-founded AutoESL (which is now Vitis HLS) which Xilinx acquired in 2010 and Neptune Design Automation (now part of Vivado) which Xilinx acquired in 2013. Falcon Computing is headquartered in Los Angeles, California. The company serves enterprise customers and academic institutions across the United States and China.

Related Contents:

For more Embedded, subscribe to Embedded’s weekly email newsletter.

The post Xilinx Acquires Falcon Computing to Bring Adaptive Computing to More Developers appeared first on Embedded.com.

>>>


Losing Vitality

2020-12-02 15:30 ElectronicsWeekly David Manners

China is trying to do something unique and very difficult – to build a chip industry without free speech. Stifling free speech is not something intrinsic to Chinese society. In 2003, Hu Jintao, who was president of China from 2002 to 2012, told a conference: “The removal of restrictions on the press, and the opening ...

This story continues at Losing Vitality

Or just read more coverage at Electronics Weekly

>>>


UK made: Cambridge Led buys Seward safety testers for production

2020-12-02 13:00 ElectronicsWeekly Steve Bush

UK lighting manufacturer Cambridge Led has bought two multi-function LED testers from Seaward of County Durham. A subsidiary of lighting distributor Net Led, Cambridge operates from a 36,000 sq ft factory in Cambridge and has been designing, manufacturing and supplying LED luminaires and systems for installation in commercial and industrial areas across the UK for over ...

This story continues at UK made: Cambridge Led buys Seward safety testers for production

Or just read more coverage at Electronics Weekly

>>>


Protection diodes for automotive CAN-FD

2020-12-02 12:21 ElectronicsWeekly Steve Bush

Nexperia has announced lead-less ESD protection devices for automotive CAN-FD applications, qualified to AEC-Q101 They are the: PESD2CANFDxxx-QC – packaged in 1.4 x 1.2mm DFN1412D-3 (SOT8009) PESD2CANFDxxx-QB – packaged in 1.1 x 1mm DFN1110D-3 (SOT8015) The 0.48mm high packages have side-wettable flanks for automated optical inspection during production, and have smaller footprints than SOT23 or SOT323 packages. Clamping ...

This story continues at Protection diodes for automotive CAN-FD

Or just read more coverage at Electronics Weekly

>>>


High CRI white leds for automotive internal lighting

2020-12-02 11:15 ElectronicsWeekly Steve Bush

Osram has introduced two ranges of AEC-Q102 qualified high-CRI white leds for automotive interiors. Ostune E1608 1.6 x 0.8 x 0.6mm (real base part number KW DELSS2.CC) 1,800 – 4,500mcd (30mA 3.2Vf typ) Ostune E3030 3 x 3 x 0.65mm (real base part number KW DSLP31.CC 18 – 33 lm (60mA (2.9Vf typ) products – KW DSLP31.CC-G) 45 ...

This story continues at High CRI white leds for automotive internal lighting

Or just read more coverage at Electronics Weekly

>>>


Video: LMD measuring your blood pressure with your smartphone

2020-12-02 11:03 ElectronicsWeekly Alun Williams

Leman Micro Devices (LMD) demonstrates its V-Sensor for smartphones and wearables, for measuring blood pressure and other vital signs

This story continues at Video: LMD measuring your blood pressure with your smartphone

Or just read more coverage at Electronics Weekly

>>>


Q3 smartphones up 20% q-o-q

2020-12-02 10:46 ElectronicsWeekly David Manners

Q3 smartphone shipments rose 20% q-o-q to 336 million units, says TrendForce. Q4=production is estimated to reach 351 million units, showing a QoQ increase of 4%. Samsung and Xiaomi were the only companies out of the top six to increase their market shares in 3Q20 while Apple fell to fourth place due to the delayed ...

This story continues at Q3 smartphones up 20% q-o-q

Or just read more coverage at Electronics Weekly

>>>


Power Profiler Kit II measures power consumption in Nordic Semi based embedded systems

2020-12-02 10:37 CNXSoft Jean-Luc Aufranc (CNXSoft)

If you’re developing battery-powered products it’s important to optimize your application to consume as little...

>>>


Khadas Tone 2 Pro mini desktop Hi-Fi system supports MQA decoding, balanced RCA outputs

2020-12-02 08:18 CNXSoft Jean-Luc Aufranc (CNXSoft)

Khadas Tone high-resolution audio board was introduced in 2018 for audiophiles wanting to design their...

>>>


Paragraf and RollsRoyce commercialise graphene Hall-Effect sensors

2020-12-02 07:28 ElectronicsWeekly David Manners

Paragraf, the graphene specialist, is setting up a supply chain for graphene Hall-Effect sensors used in high-temperature Power Electronics, Electric Machines and Drives (PEMD) within the aerospace sector. Named High-T Hall, the project stems from the UK Research and Innovation’s (UKRI) ‘Driving the Electric Revolution’ challenge and brings together Paragraf, Rolls-Royce, TT Electronics (Aero Stanrew) ...

This story continues at Paragraf and RollsRoyce commercialise graphene Hall-Effect sensors

Or just read more coverage at Electronics Weekly

>>>


Xilinx acquires Falcon Computing

2020-12-02 07:23 ElectronicsWeekly David Manners

Xilinx has acquired Falcon Computing Solutions, a privately-held provider of high-level synthesis (HLS) compiler optimization technology for hardware acceleration of software applications. The acquisition will make adaptive computing more accessible to software developers by enhancing the Xilinx Vitis Unified Software Platform with automated hardware-aware optimizations. The integration of Falcon Computing’s innovative compiler technology into the ...

This story continues at Xilinx acquires Falcon Computing

Or just read more coverage at Electronics Weekly

>>>