Sony’s Post

“The reality is, everything that moves will be autonomous,” said Danny Shapiro from NVIDIA about the inevitability of self-driving technology ➡https://lnkd.in/eb6ajhgn I recently sat down with Danny to discuss the tech giant's monumental role in crafting the cars of the future and how their Blackwell AI chip, is going to change the game for autonomous vehicles ⚡ ➡Watch the full interview HERE https://lnkd.in/eb6ajhgn to hear more about the company’s revolutionary tech and what the future of autonomy could look like. #nvidia #aichip #blackwell #blackwellchip #selfdrivingcars #nvidiaai #electricvehicles #mobility #evcharging #smartcities #evbatteries #sustainabletransport #electriccars #electrification #charginginfrastructure #autonomousvehicles #microgrids #fleets

I’m thrilled to announce that our latest research paper, co-authored with PhD student Yunlong Wang, has just been published! We’ve tackled a critical issue in the realm of Shared Autonomous Vehicles (SAVs) and their potential to revolutionize urban transportation. In our paper, we delve into how SAVs, paired with Dynamic Wireless Charging (DWC) technology, can operate continuously to meet passenger demands. This addresses the significant challenge of routing SAVs efficiently while keeping them charged on the go. Traditional methods often rely on rule-based algorithms that fall short in dealing with the complex needs of this problem. To overcome this, we propose a novel approach: the Deep Slack Induction by String Removals-based Reinforcement Learning (DSRL) framework. In simpler terms, this framework optimizes the routes of SAVs to be cost-effective and energy-efficient, ensuring they can run smoothly without any stops for charging - in other words - completely charging on the go! Here’s why this matters: • Cost-Efficiency: Our framework significantly reduces the operational costs of SAVs. • Energy Stability: It maintains a balanced charge level in SAVs, preventing unexpected downtimes. • Continuous Operation: Ensures that SAVs can meet passenger demands around the clock. Our experiments show that DSRL not only outperforms traditional methods but also adapts well to different conditions and scales, thanks to its robust learning capabilities. This means more reliable and efficient SAV services for urban commuters. Additionally, we’ve conducted sensitivity analysis that provides valuable insights for optimizing DWC infrastructure, helping city planners and engineers design better road layouts and charging solutions. For more details, feel free to reach out or check out the paper here. Which is currently in Early Access at the IEEE Transactions on Intelligent Transport Systems (a leading journal in transport) https://lnkd.in/gySzKVvP #Research #UrbanTransportation #AutonomousVehicles #WirelessCharging #SustainableMobility #Innovation

CanEduDev Partners with KTH and Telenor on Advanced Autonomous Driving Project Project Overview: CanEduDev, in collaboration with the Royal Institute of Technology (KTH) and Telenor, has initiated the project "Autonomous Drive of a Minicar Using Edge Computing and 5G Communication." This project is a part of the CART (Connected, Collaborating, and Automated Road Traffic) subproject under TECoSA (Center for Trustworthy Edge Computing Systems and Applications) and aims to develop a low-cost and safe verifier for future full-size autonomous vehicle testbeds using the CanEduDev Rover platform. Key Participants: CanEduDev: Supplying the CanEduDev Rover platform and contributing expertise in autonomous vehicle technology. KTH: Providing and integrating perception sensors and software to enable autonomous driving functionalities in the CanEduDev Rover. Telenor: Offering 5G communication services and edge servers to support real-time data processing and decision-making. Project Objectives: The project is designed to demonstrate the capabilities of edge computing and 5G communication in enhancing autonomous vehicle performance. By offloading computationally intensive tasks to nearby edge servers, the project aims to address the limitations of onboard microcontrollers in collecting necessary data and computing optimal, safe driving decisions. Key research questions include: Evaluating offloading strategies for computing tasks. Ensuring end-to-end latency meets real-time requirements for safety-critical functions. Evaluating the impact of edge computing architecture on vehicle safety and performance. Significance: This initiative marks the beginning of developing a safe, secure, and effective R&D platform for the industry. By leveraging emerging technologies, the project's CanEduDev Rover platform—equipped with advanced sensors and communication systems—will serve as a scalable, safe, and cost-effective solution for testing and validating new technologies. Implications for CanEduDev: This project aligns with CanEduDev's strategic objective of creating an effective R&D platform for innovation, higher education, and research. It represents a significant step towards developing new technological capabilities for its scale model cars and achieving the long-term goal of advancing cutting-edge technology solutions through partnerships and innovative research. Tags: #Innovation #EdgeComputing #5G #AutonomousVehicles #RoverPlatform #CanEduDev #KTH #Telenor #ResearchAndDevelopment #HigherEducation #Technology #SmartTransportation #Collaboration #TechResearch #R&D

Navigating the intricate landscape of autonomous vehicle (AV) development, engineers grapple with the demanding requirements of reliability and safety, distinct from consumer electronics. Faction, a company at the forefront of AV autonomy, strategically partners with Point One Navigation, a Hemi portfolio, to address these challenges. Point One provides a uniform API that works across the journey from the large-scale Atlas boxes to integrated software that runs alongside the “chip down” designs. This has allowed Faction to scale its range of vehicle platforms without changing the embedded code, saving months of retest and validation time. This collaboration optimizes scaling efforts, reducing typical hardware redesign delays inherent in AV development. #HemiAuto Aaron Nathan https://lnkd.in/gzGvnuNQ

Virtual Sensors are revolutionizing the way we approach complex system development, making it more efficient and cost-effective. From autonomous cars to #AR, discover how digital twins are making a significant impact. Read this article: https://lnkd.in/ggY7CN94 #digitaltwin #autonomousdriving #lidar #ar #virtualsensors

Exploring the future of Self-Driving Vehicles (SDVs) opens a gateway to unprecedented opportunities. From reshaping urban mobility to advancing safety standards, this transformative technology heralds a new era. Join us as we navigate through the intricacies of SDVs, unraveling the potential they hold and the promising horizons they unveil. The road ahead is not just autonomous; it's a journey into a future brimming with possibilities and advancements. #BITSILICA #VLSI #Semiconductors #Embedded #AI #VLSIJobs #Technology #CareerGrowth #DigitalIndia #SelfDrivingVehicles #tech #technology #technews #gadgets #innovation #business #techno #programming

Taking the next steps with CanEduDev ;)

As autonomous driving becomes a reality, manufacturers are looking for solutions that can train systems to drive autonomously without the need for human intervention. This project aims to achieve consistency in multi-object tracking across different timescales. The end result was a self-driving car that could handle multiple objects with ease. This is great news. #AutomotiveIndustry #AutonomousDriving #SelfDrivingCar https://lnkd.in/eVskYRmU

Ambarella and Plus announce that PlusVision™—a high-performance transformer-based autonomous driving perception software stack with industry-leading power efficiency—is now available on Ambarella’s CV3-AD AI domain controller family of systems-on-chip (SoCs), which provides industry leading AI performance per watt. This cutting edge birds-eye-view vision technology will help Tier-1 suppliers and carmakers to accelerate the launch of next-generation vehicles with L2+/L3 autonomy. Additionally, it enables them to create a future-proof AD platform to efficiently differentiate their offerings and maximize customer value. Read today’s press release, here: https://ow.ly/6pZL50SNTBP #Ambarella #Plus #AIenvisioned #AI #Automotive #AutonomousVehicles #Innovation #Technology #TransformerNetworks #ADSoftwareStack #BirdsEyeView #AutonomousDriving #AIDomainController #AISoC #Autonomy

[COMST Survey] Beam Alignment in MmWave V2X Communications: A Survey The digital transformation within the automotive industry is accelerating towards an era dominated by autonomous vehicles, with vehicle-to-everything (V2X) communications being a fundamental enabler for this advancement. As vehicular networks evolve to meet the complex demands of autonomous driving, traditional communication systems encounter limitations in bandwidth and data transfer rates. Millimeter-wave (mmWave) communication emerges as a pivotal solution, offering the extensive bandwidth required for the high data throughput and low latency essential in modern vehicular communications. However, challenges loom, with beam alignment in mmWave V2X becoming a time-consuming process and the mmWave’s blockage effect impeding consistent and reliable vehicular communication links. Therefore, the development of efficient, real-time, and robust beam alignment technology is crucial for mmWave V2X communication. In this context, the paper titled “Beam Alignment in MmWave V2X Communications: A Survey” by Jingru Tan, Tom H. Luan, Wenbo Guan, Yuntao Wang, Haixia Peng, Yao Zhang, Dongmei Zhao and Ning Lu presents a survey. First, the paper explores various approaches including beam sweeping, angle of arrival (AoA)/angle of direction (AoD) estimation, black-box optimization, and side information. Then, the paper introduces performance metrics for assessing beam alignment performance and compare the performance of four beam alignment methods under different metrics. Finally, the paper summarizes the future research directions and challenges faced by beam alignment techniques in mmWave V2X communication.