Search for: All records

Path planning is a critical task for autonomous driving, aiming to generate smooth, collision-free, and feasible paths based on input perception and localization information. The planning task is both highly time-sensitive and computationally intensive, posing significant challenges to resource-constrained autonomous driving hardware. In this article, we propose an end-to-end framework for accelerating path planning on FPGA platforms. This framework focuses on accelerating quadratic programming (QP) solving, which is the core of optimization-based path planning and has the most computationally-intensive workloads. Our method leverages a hardware-friendly alternating direction method of multipliers (ADMM) to solve QP problems while employing a highly parallelizable preconditioned conjugate gradient (PCG) method for solving the associated linear systems. We analyze the sparse patterns of matrix operations in QP and design customized storage schemes along with efficient sparse matrix multiplication and sparse matrix-vector multiplication units. Our customized design significantly reduces resource consumption for data storage and computation while dramatically speeding up matrix operations. Additionally, we propose a multi-level dataflow optimization strategy. Within individual operators, we achieve acceleration through parallelization and pipelining. For different operators in an algorithm, we analyze inter-operator data dependencies to enable fine-grained pipelining. At the system level, we map different steps of the planning process to the CPU and FPGA and pipeline these steps to enhance end-to-end throughput. We implement and validate our design on the AMD ZCU102 platform. Our implementation achieves state-of-the-art performance in both latency and energy efficiency compared with existing works, including an average 1.48× speedup over the best FPGA-based design, a 2.89× speedup compared with the state-of-the-art QP solver on an Intel i7-11800H CPU, a 5.62× speedup over an ARM Cortex-A57 embedded CPU, and a 1.56× speedup over state-of-the-art GPU-based work. Furthermore, our design delivers a 2.05× improvement in throughput compared with the state-of-the-art FPGA-based design. 

In this article, we share our real-world experiences of digital twin, a practical autonomous driving system development paradigm, which generates an integral, comprehensive, precise, and reliable representation of the physical environment to minimize the need for physical testing. 

In many tissues, cell type varies over single-cell length-scales, creating detailed heterogeneities fundamental to physiological function. To gain understanding of the relationship between tissue function and detailed structure, and eventually to engineer structurally and physiologically accurate tissues, we need the ability to assemble 3D cellular structures having the level of detail found in living tissue. Here we introduce a method of 3D cell assembly having a level of precision finer than the single-cell scale. With this method we create detailed cellular patterns, demonstrating that cell type can be varied over the single-cell scale and showing function after their assembly. 

Autonomous Driving (AD) is a rapidly developing technology and its security issues have been studied by various recent research works. With the growing interest and investment in leveraging intelligent infrastructure support for practical AD, AD system may have new opportunities to defend against existing AD attacks. In this paper, we are the first to systematically explore such a new AD security design space leveraging emerging infrastructure-side support, which we call Infrastructure-Aided Autonomous Driving Defense (I-A2D2). We first taxonomize existing AD attacks based on infrastructure-side capabilities, and then analyze potential I-A2D2 design opportunities and requirements. We further discuss the potential design challenges for these I-A2D2 design directions to be effective in practice. 

Abstract A new concentrated ternary salt ether‐based electrolyte enables stable cycling of lithium metal battery (LMB) cells with high‐mass‐loading (13.8 mg cm⁻², 2.5 mAh cm⁻²) NMC622 (LiNi_0.6Co_0.2Mn_0.2O₂) cathodes and 50 μm Li anodes. Termed “CETHER‐3,” this electrolyte is based on LiTFSI, LiDFOB, and LiBF₄with 5 vol% fluorinated ethylene carbonate in 1,2‐dimethoxyethane. Commercial carbonate and state‐of‐the‐art binary salt ether electrolytes were also tested as baselines. With CETHER‐3, the electrochemical performance of the full‐cell battery is among the most favorably reported in terms of high‐voltage cycling stability. For example, LiNi_xMn_yCo_1–x–yO₂(NMC)‐Li metal cells retain 80% capacity at 430 cycles with a 4.4 V cut‐off and 83% capacity at 100 cycles with a 4.5 V cut‐off (charge at C/5, discharge at C/2). According to simulation by density functional theory and molecular dynamics, this favorable performance is an outcome of enhanced coordination between Li⁺and the solvent/salt molecules. Combining advanced microscopy (high‐resolution transmission electron microscopy, scanning electron microscopy) and surface science (X‐ray photoelectron spectroscopy, time‐of‐fight secondary ion mass spectroscopy, Fourier‐transform infrared spectroscopy, Raman spectroscopy), it is demonstrated that a thinner and more stable cathode electrolyte interphase (CEI) and solid electrolyte interphase (SEI) are formed. The CEI is rich in lithium sulfide (Li₂SO₃), while the SEI is rich in Li₃N and LiF. During cycling, the CEI/SEI suppresses both the deleterious transformation of the cathode R‐3m layered near‐surface structure into disordered rock salt and the growth of lithium metal dendrites.