Smart Cities Archives - Page 364 of 570 - SmartCitiesElectronics.com

Scalable AI SMARC SoM combo targets HMI and embedded vision

Thu, Nov 12, 2020

Comprising 10 Renesas ICs, the scalable system-on-module (SoM) Smart Mobility ARChitecture (SMARC) Winning Combination board accelerates development of artificial intelligence (AI) IoT face/object detection, image processing, and 4K video playback applications, says Renesas. The board can be used in surveillance cameras, inspection equipment, and industrial and building automation HMI and embedded vision systems.

The 10 ICs include microprocessor (MPU), power and analog ICs. The board is based on the SMARC 2.0 industry standard, which specifies an 82 x 80mm form factor. The SMARC SoM board offers designers a choice of three different scalable versions of the Renesas 64-bit RZ/G2 MPU. Options are an RZ/G2N dual core Arm Cortex-A57 MPU operating at 1.5 GHz for mid-range performance; the RZ/G2M MPU with dual-core Arm Cortex-A57 and quad-core Arm Cortex-A53 (1.2 GHz) for high performance; and the RZ/G2H MPU with quad-core Arm Cortex-A57 and quad-core Arm Cortex-A53 for what Renesas describes as ultra-high performance. All three MPUs (two cores up to eight cores) feature integrated 600MHz PowerVR 3D graphics and a 4K UHD H.265 and H.264 codecs to satisfy the needs of different computer processing requirements.

DK Singh, director, of the Systems and Solutions team at Renesas, said: “Our scalable solution’s extensive on-board interfaces, large memory and programmable clocks for different needs makes it an excellent turnkey solution for a wide range of applications. And its industry leading RZ/G2 MPUs with 3D graphics engine and 4K UHD and full HD video codec provides higher performance per dollar than competing 64-bit MPUs and GPUs.”

The SMARC SoM offers designers their choice of RZ/G2 MPU with up to 35.6k DMIPS performance, between 2.0 and 4Gbyte LPDDR4 RAM and 32Gbyte eMMC. Each RZ/G2 MPU is capable of running edge video analytics and AI frameworks. The MPUs feature an integrated AI software library, a set of interfaces, error checking and correction (ECC) protection on both internal and external memories, Linux OS, and a Verified Linux Package (VLP) which is tested and maintained by Renesas. There is also a Civil Infrastructure Platform (CIP) Super Long-Term Support (SLTS) kernel, and a Linux kernel bundled with a software development environment. The SMARC SoM board provides an optimised power and programmable timing tree. There is also support for dual-band Wi-Fi and Bluetooth Low Energy (BLE) for wireless communication, fast communication interfaces such as USB, SATA, LVDS, HDMI, CSI, I2S, PCIe, and Gigabit Ethernet, all accessible through the SMARC 2.0 connector.

The board includes the ISL1208 low-power real time clock for calendar-based applications powered by a small 400nA battery, or supercapacitor during a power failure and the P8330 power management IC (PMIC) for delivering power to multiple supply rails.

Also included are two clock sources, Renesas’ small form factor VersaClock 3S programmable clocks with integrated 32.768kHz DCO powered by a single coin cell battery during a power failure and a PCIe clock generated with the Renesas 9FGV0641. This supports six 100MHz differential clock outputs, with PCIe Gen 1 to 4 support

There is also the option for connection to two external cameras and LCD panel with capacitive touch.

The Scalable AI SMARC SoM Winning Combo Solution board was designed by Renesas and developed through a collaboration with RelySys Technologies. The board and design files are available now.

http://www.renesas.com

> Read More

Reference design for ToF lidar supports range finding for precision robotics

Tue, Nov 10, 2020

Light detection and ranging (lidar) is increasingly popular in robotics and industrial proximity sensing where mm range accuracy is required. ON Semiconductor has introduced the a reference design based on its Silicon Photomultiplier (SiPM) technology.

The SiPM direct time of flight (dToF) lidar platform exploits the dToF method, which measures the time it takes for a pulse of light, normally in the near infrared (NIR) wavelength range, to travel to and from an object. This deceptively simple technology can incur challenges, such as high levels of ambient solar light. To determine range accurately, the receiver needs to capture as much of the signal as possible. ON Semiconductor says that the SiPM sensor overcomes the shortfall of traditional photodiodes by providing faster response times and high detection efficiency. The reference platform uses the RB-Series, second generation of SiPM sensors, which deliver improved performance in red and the NIR range.

The reference design includes the NIR laser diode, SiPM sensor and optics, as well as the digital processing necessary to convert the detected signals into elapsed time, and elapsed time into distance. OEMs can adapt it for production of industrial range finding applications.

To accelerate customers’ time to market, ON Semiconductor has made all of the design data for the reference platform available, covering the schematics, BoM, Gerber files, and PCB design files. A PC-based GUI is also accessible, which provides a graphical representation of the measurements over time. The histograms generated provide further evidence of the system’s capabilities in applications such as range finding, collision detection and 3D mapping.

The SiPM dToF lidar platform can detect objects at distances between 100mm and 23m. The design is FDA Class 1 certified and compliant with IEC / EN 60825-1:2014 and 21 CFR 1040.10/ 1040.11 laser safety standards.

ON Semiconductor supplies semiconductor-based solutions, offering a portfolio of energy efficient, power management, analogue, sensors, logic, timing, connectivity, discrete, SoC and custom devices for automotive, communications, computing, consumer, industrial, medical, aerospace and defence applications.

ON Semiconductor operates a network of manufacturing facilities, sales offices and design centres in key markets throughout North America, Europe and the Asia Pacific regions.

http://www.onsemi.com

> Read More

Automotive GaN FETs can double power density, says Texas Instruments

Tue, Nov 10, 2020

Adding to its high voltage power management portfolio, Texas Instruments announces 600 and 650V GaN FETs for automotive and industrial applications. They have a fast switching, 2.2MHz integrated gate driver, which can allow designers to deliver twice the power density, achieve 99 per cent efficiency and reduce the size of power magnetics by 59 per cent compared to existing solutions, says TI.

The LMG3422R050, LMG3425R050, LMG3422R030 and LMG3425R030 FETs use its proprietary GaN materials and processing capabilities on a GaN-on-silicon (Si) substrate, which provides a cost and supply chain advantage over comparable substrate materials such as silicon carbide (SiC), maintains TI.

TI says the automotive GaN FETs can help reduce the size of electric vehicle (EV) onboard chargers and DC/DC converters by as much as 50 per cent compared to existing Si or SiC solutions. Including them in vehicles can extend battery range, increase system reliability and lower design cost to meet the increased desire for electrification in vehicles.

In industrial designs, the devices enable high efficiency and power density in AC/DC power delivery applications where low losses and reduced board space are important, for example in hyperscale and enterprise computing and 5G telecomms rectifiers.

As electronic systems are getting smaller, the components used must get smaller and sit closer together, explains TI. The GaN FETs integrate a fast switching driver, internal protection and temperature sensing, to reduce board space in high performance power management designs. The company says that this level of integration, coupled with the high power density of TI’s GaN technology, enables engineers to eliminate more than 10 components typically required for discrete solutions. Furthermore, each of the new 30mOhm FETs can support up to 4kW of power conversion when applied in a half bridge configuration.

To address the challenge of fast switching capability at the cost of higher power losses, the GaN FETs feature TI’s ideal diode mode to reduce power losses. TI reports, for example, in PFCs, ideal diode mode reduces third-quadrant losses by up to 66 per cent compared to discrete GaN and SiC MOSFETs. Ideal diode mode also eliminates the need for adaptive deadtime control, reducing firmware complexity and development time.

Offering 23 per cent lower thermal impedance than the nearest competitive packaging, the TI GaN FET packaging allows engineers to use smaller heat sinks while simplifying thermal designs. They also provide thermal design flexibility, with the ability to choose from either a bottom- or top-side-cooled package. The FETs’ integrated digital temperature reporting enables active power management, for engineers to optimise system thermal performance under varying loads and operating conditions.

Pre-production versions of the four industrial-grade, 600V GaN FETs are available now in a 12 x 12mm, quad flat no-lead (QFN) package. Volume production is expected to begin in Q1 2021.

Evaluation modules are each FET available now.

Pre-production versions of the LMG3522R030-Q1 and LMG3525R030-Q1 650V automotive GaN FETs and evaluation modules are also scheduled for Q1 2021. Engineering samples are available upon request.

http://www.TI.com

> Read More

EdgeReady provides touchless access control via face recognition

Mon, Nov 09, 2020

Based on its i.MX RT106F crossover microcontroller, NXP offers EdgeReady for the IoT, with a license for facial recognition software to provide touchless access control.

OEMs can use EdgeReady to quickly, easily and inexpensively add vision-based touchless access control, says NXP. The microcontroller and AI/ML-based face recognition software makes possible face recognition with liveness checking using infra red and RGB cameras, flash and SDRAM at a total cost of less than US$10, says the company.

It can be used to implement face recognition for authenticating and operating secure machines, equipment, and access points for hotels, public buildings and homes, for example, doors, locks, amenities, lifts, garage doors and security systems.

The infra red camera operates with a visible light (RGB) camera to distinguish a real person from a photograph. This prevents unauthorised access without the need for an expensive 3D structured light or time of flight (ToF) camera, says NXP.

It responds in 0.5 seconds. Face detection, quality check, liveness detection and face recognition are all conducted in less than 500 milliseconds at the microcontroller-based edge instead of the cloud, to make access systems convenient.

It also enhances data privacy by allowing all personal biometric data to remain on the device at the edge. This helps address privacy regulations as well as the concerns of consumers to protect data.

The battery life is maximised with a fast microcontroller boot up time. The device enters power saving mode when not in use. A fast recognition time is combined with the microcontrollers low active power, says NXP, to further meet low power system requirements.

A customisable smartphone or PC application can enable biometric models to be sent to remote locks and other devices for controlled access in hospitality, smart building and smart home applications.

NXP also offers the i.MX RT106F-based development kit (SLN-VIZNAS-IOT) to accelerate time to market. It contains all the software and hardware information to complete designs in as little as six months, says NXP. In addition to negating integration and coding, the NXP EdgeReady IoT for secure face recognition based on the i.MX RT106F development kit includes schematics, bill of materials (BoM), board layout and ready-to-use face biometric registration application.

The SLN-VIZNAS-IOT secure face recognition development kit is available now from NXP and authorised distributors.

The i.MX RT106F microcontroller is available now. It includes a license to use the NXP face recognition SDK. The device is available in consumer and industrial temperature grades.

http://www.nxp.com

> Read More

Scalable AI SMARC SoM combo targets HMI and embedded vision

Reference design for ToF lidar supports range finding for precision robotics

Automotive GaN FETs can double power density, says Texas Instruments

EdgeReady provides touchless access control via face recognition

About Smart Cities