DFI focuses on AI edge for smart transportation

At Computex Taipei this week, embedded motherboard and industrial computer specialist, DFI has introduced its low power vehicle system, based on Arm architecture.

The high performance computing micro-industrial motherboard is compact, allowing it to overcome space limitations in roadside equipment. It can be used for people flow identification in vehicle systems. DFI said it can improve energy efficiency and provide safety for future cities through intelligent, automated and miniaturised product design.

Exhibiting in the BenQ Group Computex Smart+ exhibition stand number MO619a at Computex Taipei (30 May to 2 June) at the Taipei Nangang Exhibition Center, Taipei, Taiwan, DFI has actively deployed smart transportation solutions to upgrade public infrastructure, lower operating costs and increase efficiency, said DFI president, Alexander Su.

DFI’s smart transportation embedded solutions support road and information security. The company has joined VicOne, a vehicle information security company and IoT system integration company, 5GIoTLead Technology, to introduce in-vehicle cybersecurity technology and the 5G Smart Pole Management platform, which demonstrate simulations of smart traffic intersections.

Another highlight is the miniature industrial fanless EC70A-TGU, equipped with the 11th generation Intel Tiger Lake Core processor. This processor has low power consumption and built-in GPU based on the Intel Iris Xe architecture, for edge computing. The fanless design allows the EC70A-TGU to be light and compact. Its accurate AI judgment also considers performance and real-time responses, making it ideal for handling heavy workloads, said DFI.

The EC70A-TGU can perform real-time AI calculations and speed up pedestrian recognition, tracking, and video streaming. It can also identify violations committed by others and synchronously send the information to the smart pole management platform, providing a real-time visual dashboard and data control to enhance management efficiency.

Another innovation is the VC900-M8M in-vehicle system which adopts low power ARM architecture and supports in-vehicle information security software.

The VC900-M8M, with ARM-based NXP i.MX8M CPU, can support Yocto Linux 2.5 and Android 9.0 operating systems. It has a variety of I/O configurations and features an anti-vibration design and wide operating temperature range to allow the vehicle to function even in extreme environments, said DFI. 

The VC900-M8M has a built-in six-axis IMU sensor, which monitors vehicle and driver behaviour to prevent speeding, sudden braking, and collisions. It can also manage abnormal signals and provide real-time warnings.

https://www.dfi.com/

> Read More

Compact, low power RF sensors detect position and passenger movement

RF ranging sensors for automotive applications in the SC1260 series use the 60GHz frequency band. Socionext has also integrated time division multiplexing (TDM-MIMO) operation with multiple transmitting and receiving antennas to detect the position and movement of passengers in a vehicle with maximum accuracy.

The SC1260 incorporates antennas and radar signal operation circuits for maximum detection accuracy in a low power, compact package. The sensors use a wide band of 6.8GHz (57.1 to 63.9GHz), TDM-MIMO and built-in radar signal operation circuits for range-finding and angle calculation.

TDM-MIMO operation enables high-precision sensing, such as the detection of passengers when three people are sitting in a row inside the vehicle, added Socionext.

The sensors are claimed to enable easy acquisition of 3D position information without requiring advanced expertise in high frequency devices and signal operation. Each sensor combines antennas, RF circuits, ADCs, FIFO memory, an SPI and intelligent power control sequencer for flexible duty cycle control. The sensor requires only 0.72mW at 0.1 per cent duty cycle operation.

The sensors are suitable for a range of automotive applications, such as tracking the position and movements of passengers in a vehicle while suppressing the load on the vehicle battery when the engine is off, said Socionext.

The SC1260 series complies with the global broadband 60GHz radio equipment standard. Sample and evaluation kit shipments are scheduled for June, with production volumes to be available in Q1 2024.

The company is believed to be the first to develop compact, low power 60GHz RF ranging sensors with built-in radar signal operation circuits for range-finding and angle calculations. It used its experience and knowledge accumulated through the development of millimeter-frequency wireless communication ICS and 24GHz radio-frequency ranging sensors, observed Teruaki Hasegawa, head of Socionext’s IoT & radar sensing business unit.

The sensor is supplied in a FC-BGA with dimensions of 6.0 x 9.0 x 1.2mm.

Socionext Europe is a fabless SoC supplier. Its headquarters is in Frankfurt, Germany with offices in Munich, Germany and Maidenhead, UK. 

The Munich Design and Support Centre, specialises in hardware and software product development for the automotive sector. Maidenhead supports Socionext’s IP Development & Engineering Centre with a focus on high speed data conversion IP. 

https://www.socionext.com/en/

> Read More

Edge sensors in AMD’s Artix UltraScale+ family are ADAS-optimised

Addressing market needs for faster signal processing and smaller form factors in systems using edge sensors, such as lidar, radar and 3D surround-view camera systems, AMD has added two edge sensors to its Automotive XA Artix UltraScale+ family.

The XA AU10P and XA AU15P processors are cost-optimised processors, which is a requirement in the consumer driven automotive market. They are automotive-qualified and optimised for use in advanced driver assistance systems (ADAS) sensor applications. The Artix UltraScale+ devices join the automotive-grade Spartan 7, Zynq 7000 and Zynq UltraScale+ product families. 

The XA Artix FPGAs are shipping now. They offer high serial bandwidth and signal compute density in an compact form factor. The Artix UltraScale+ devices maximise system performance via DSP bandwidth for cost-sensitive and low-power ADAS edge applications including networking, vision and video processing and secure connectivity. 

XA Artix UltraScale+ processors are certified for functional safety up to ASIL-B, which is critical for automotive ADAS sensors from cameras to lidar systems, said AMD, for use in vehicles and robotaxis, for example.

“As automotive systems have grown in complexity, safety is more critical than ever before with automotive OEMs and Tier 1 suppliers requiring ASIL-B certification for LiDAR, radar and smart edge sensor applications,” said Ian Riches, vice president, global automotive practice, TechInsights. 

Automotive designers can use XA Artix UltraScale+ devices for sensor fusion, bringing in data from multiple edge sensors and performing image and video processing before porting it to an external SoC. The new XA Artix UltraScale+ devices can also be connected to multiple displays in the vehicle to enhance infotainment features, advised AMD.

http://www.xilinx.com 

> Read More

Battery-operated video camera systems add AI for cloud IoT devices

InnoPhase IoT adds AI and solar panel augmentation to battery-operated video camera system via the Talaria TWO ULP (ultra low power) Wi-fi-optimised reference kit for cloud-connected IP video IoT devices. 

In addition to applications such as battery-operated video cameras, smart video doorbells, wearables, smart appliances, home security cameras and in-vehicle dashboard monitoring devices, OEMs and ODMs are introducing new categories of video devices with designs that use solar technology and AI, said InnoPhase IoT.

Based on the Talaria TWO ULP Wi-Fi and Bluetooth Low Energy (BLE) platform, the reference kit consists of dual stack, power optimised software combined with an advanced digital polar radio design to address the challenges involved with yesterday’s power-hungry processing of video IoT designs with a multi-protocol, cloud connected camera system, said the company. 

The InnoPhase IoT Talaria TWO wireless platform for video supports up to 2K camera resolution and has an integrated microcontroller which enables off-loading of TCP-IP networking and cloud connectivity stack when ISP (image signal processing) is powered down. An always on, always connected enables low latency and mitigates image loss issues and the low power Wi-Fi at BLE power levels enables a two to four times improvement in battery life.

Until now, untethered wireless cloud connectivity and a long battery life were major design barriers for video IoT. Approximately 10 per cent of video cameras are currently battery-operated primarily because of battery life limitations, reported InnoPhase IoT. Moving to a wireless format wasn’t possible since ubiquitous Wi-Fi is inherently power hungry, rapidly draining video camera’s batteries. The possibility of adding even more features such as AI would mean an even shorter battery life.

The new InnoPhase IoT video reference kit address all of these challenges and provides 40 per cent lower power consumption and a battery life in excess of one year. The company also said it offers device developers a variety of market ready solutions from OEMs and ODMs.

“Using InnoPhase IoT Talaria TWO and an Ingenic T31 integrated reference design, we have been able to build an AI-enabled smart video camera with battery life two to three times longer than today’s solutions. We’re also achieving multi-year battery life when augmented with a solar panel”, said Larry Yang, product manager of smart product line at CE-Link, a wireless video camera OEM. The reference design can be customised, enabling customers to quickly bring branded wireless camera systems to market, he added.

According to Jason Lim, CEO at Kenxen, Taiwan, a smart video camera ODM, the Talaria TWO Wi-Fi / BLE module enables low power, direct-to-cloud connectivity for battery operated cameras. “This allows us to offer quick time to market video products for our end customers,” he added.

InnoPhase IoT president and COO, Wiren Perera, said: “InnoPhase IoT’s Talaria TWO untethers cameras from all wired power and network connections, resulting in more accurate data for cloud processing. It supports the addition of enhanced features such as artificial intelligence and unleashes design creativity only possible with the low-power, long battery-life model.”

The reference kits are currently available for evaluation. Each includes hardware and software integration with an image signal processor, cloud connectivity software and access to ODMs.

http://www.innophaseiot.com 

> Read More

About Smart Cities

This news story is brought to you by smartcitieselectronics.com, the specialist site dedicated to delivering information about what’s new in the Smart City Electronics industry, with daily news updates, new products and industry news. To stay up-to-date, register to receive our weekly newsletters and keep yourself informed on the latest technology news and new products from around the globe. Simply click this link to register here: Smart Cities Registration