Browse Topic: Hardware
This paper details the successful scaling demonstration of a comprehensive supply chain screening process for commercial off-the-shelf (COTS) motherboard subassemblies used in tactical servers for naval applications. Our approach leverages Power Fingerprinting (PFP) technology, which uses unintended analog emissions and machine learning to provide independent, non-destructive, and scalable integrity assessment of microelectronics. The primary goal of the effort was to demonstrate the effectiveness and scalability of the PFP screening process without disrupting or delaying the manufacturing workflow. The screening successfully detected hardware and firmware modifications and identified two cases of abnormal behavior: unusual BIOS power reset and elevated CPU sensor readings on two motherboard subassemblies. Following our quality control forensic analysis, we determined the root cause of these anomalies and their potential impact on the host platform.
It was reported earlier that the wear differential between the inboard pad and the outboard pad leads to brake squeal generation. The (inboard/outboard) pads wear differential can occur due to hardware issues such as brake pad drag and/or two different wear rates of the (I/O) pads, which is caused by two different material properties of the pads although the pad formula may be the same. It is found that (I/O) pads compressibility differential/hardness differential/friction differential are all interrelated and that they contribute to brake squeal generation in addition to the inboard pad tangential/radial taper wear. A method has been found to separate the inboard pad friction and the outboard pad friction and to estimate each friction coefficient.
Physical AI refers to applications in which AI technologies are connected to hardware that sense and execute actions in the physical world, allowing systems to autonomously act and adapt in real time. It spans automotive, robotics, industrial automation, smart infrastructure, aerospace, healthcare devices, software-defined machines, and more. Regardless of application, they have one thing in common: they must operate safely, reliably, and predictably in real world environments. Unlike purely digital AI, these systems are constrained by embedded electronics, timing, power, safety, and system-level interactions that are difficult to validate early.
Deep Reinforcement Learning (DRL) for quadrotor flight control typically relies on Domain Randomization (DR) for sim-to-real transfer, resulting in overly conservative policies that struggle with dynamic disturbances. To overcome this, we propose a novel adaptive control architecture that actively perceives and reacts to instantaneous perturbations. First, we train an optimal outer-loop policy, then replace its reliance on ground-truth disturbance data with a Residual Dynamics Predictor (RDP). The RDP estimates the external forces and moments acting on the aircraft in flight online using only the history of states and control actions. For seamless hardware transfer, we introduce a data-efficient linear calibration bridge and an online thrust correction mechanism that align the simulated latent space with reality using mere seconds of flight data. Real-world validations on a Crazyflie micro-quadrotor demonstrate that our adaptive controller significantly outperforms baselines, maintaining precise trajectory tracking under severe uncertainties including mass variations, asymmetric payloads, and dynamic slung loads.
The automotive industry is evolving from a reactive, independently self-determined approach to cybersecurity, complicated by a complex supply chain. Over time, this has resulted in a fragmented industry comprised of any number of proprietary solutions verses a standardized, regulated paradigm to facilitate a platform-oriented approach. This document, an update on collaborative work from the SAE Vehicle Electrical Hardware Security Task Force (TEVEES18B) and GlobalPlatform Automotive Task Force, outlines this transition strategy. An extensible number of additional examples of use cases of Global Platform Technologies are explored in this document.
Software-defined vehicles (SDVs) are reshaping automotive control architectures by shifting intelligence to embedded systems, where computational efficiency is paramount. This paper presents a systematic evaluation of control strategies (PID, LQR, MPC) for the classical control problem involving inverted pendulum on a cart under strict embedded constraints representative of software-defined vehicle ECUs. The objective is to evaluate and compare the performance of advanced control algorithms under varying control objectives when deployed on microcontrollers with constrained computational and memory resources, representative of the limitations encountered in embedded platforms used for SDVs. Furthermore, the study illustrates systematic optimization strategies that enable these algorithms to achieve real-time execution within such resource-constrained environments. Each control strategy is implemented with careful consideration of algorithmic complexity, real-time responsiveness, and resource utilization. Performance is evaluated across key metrics, enabling a comparative analysis that highlights trade-offs between control fidelity and hardware efficiency. By demonstrating how advanced control logic can be effectively deployed on constrained hardware, this work supports the broader goal of enabling intelligent, responsive vehicle behavior through software-centric design. The findings are particularly relevant for automotive and embedded engineers developing control systems for SDVs, where balancing performance and resource constraints is critical to achieving scalable, safe, and adaptive vehicle functionality.
The reliability of Drive Unit (DU) oil pumps is critical to the performance and safety of electric vehicles, as these pumps provide essential lubrication and thermal management. In modern EV architectures, real-time health monitoring of these pumps typically relies on indirect signals than dedicated sensing hardware, a design choice optimized for cost, weight, and system complexity. This makes early fault detection a non-trivial challenge. To address this limitation, we present a novel, data-driven anomaly detection framework that leverages large-scale customer fleet telemetry and advanced machine learning to identify incipient pump degradation that traditional diagnostic methods often fail to capture. Specifically, we develop an XGBoost regression model trained on time-series features—including commanded pump speed, oil temperature, and historical pump current—to predict expected current behavior under nominal conditions. Deviations are quantified using the Mean Absolute Percentage Error (MAPE) between predicted and actual currents, providing a continuous and interpretable measure of anomaly severity. A fully automated pipeline ingests daily telemetry, performs session segmentation, executes predictive modeling, and records anomaly outcomes in backend databases for continuous monitoring and engineering review. The proposed framework enables continuous, fleet-wide predictive maintenance of DU oil pumps. It improves early detection of degradation, reduces vehicle downtime, enhances safety, and increases customer satisfaction. More broadly, it highlights the potential of large-scale data analytics and machine learning to advance predictive maintenance and reliability in electric vehicle (EV) systems.
This paper describes the electromagnetic noise mitigation on the Maryland Tiltrotor Rig (MTR) and presents its first hover test results. The primary source of noise was found to be pulse width modulation associated with the motor controller. Due to this noise, testing was limited to unpowered, freewheeling cases. To solve the noise problem and allow powered testing, three hardware filters were integrated into the power and data systems. A complementary digital filter was also used. With the filtering solution in place, hover tests were carried out to high collectives of 30◦and blade loadings of 0.2. The test data was assessed using blade element-momentum theory predictions.
Model Based Design (MBD) uses mathematical modelling to create, test and refine systems in simulated environment, primarily applied in control system development. This paper discusses an approach to control gear shifting using shift logic on vehicle level for twin clutch transmission using prototype controller. Twin clutch transmission is a concept with two clutches, one at input end of the transmission called primary clutch and the other at output end of the transmission called secondary clutch. This concept is proposed to counter the challenges with conventional transmission which include increased gear shift time and effort in lower gears, potential rollback of vehicle in uphill condition and chance of missed shifts. The advantages of this concept include reduced gear shift effort and improved synchronizer life with potential for reducing the size of the synchro pack. This paper proposes a methodology to develop shift logic, integrate hardware with software, flashing and calibration on vehicle using prototype controller. The shift logic is implemented using state machines in ASCET. The state machine uses vehicle speed, accelerator pedal position, gear shifter input, shift maps and current gear to determine most optimal gear shift in real-time. This control logic is then converted into c-code and integrated with hardware using INTECRIO, which is a build environment where inputs and outputs are mapped accordingly for shift actuation. In the next step, .a2l and .cod files are generated, which are flashed onto rapid prototyping hardware from ETAS using INCA. The same is used to actuate shift solenoids of transmission to change the gear. Use of rapid prototyping hardware has significantly reduced the number of iterations required for integrating software with ECU from specific supplier, thus reducing overall development time. Final vehicle level calibration is done using INCA, followed by validation process to ensure optimal performance.
This paper addresses the challenge of increasing hardware complexity, long development cycles and high costs associated with integrating multiple systems. The research explores the potential of Large Language Models (LLMs) when applied as chatbots to revolutionize the design and development of automotive electrical hardware systems, encompassing areas such as convenience features, safety systems, advanced lighting, vehicle body control and modular electronic control units. A key focus is on how LLMs can automate cost-reduction design tasks, including design optimization, requirements verification and component validation, ultimately driving down expenses without compromising performance or reliability. Furthermore, the research investigates how LLMs can assist in decision-making by providing data-driven insights that inform critical design choices and facilitate enhanced team collaboration, leading to improved productivity through innovative tools and streamlined workflows. In that sense, the project’s scope includes creating a Data Warehouse with relevant design, features offering, testing and quality feedback data to train the LLMs. A research tool based on LLMs will be developed to offer optimization recommendations, such as identifying oversizing, suggesting cost reduction and hardware modularization. The tool will be seamlessly integrated into the cost-reduction engineering teams’ workflow, promoting agility and modernization. The expectation is that this research will drive innovation, enhance competitiveness and promote sustainability within the automotive sector, leading to accelerated time-to-market for new vehicle models, improved engineering efficiency in the development of electrical hardware systems and a strengthened market position for automotive OEMs. Ultimately, the goal is to create more efficient, safer and feature-rich automotive experiences while simultaneously optimizing cost-effectiveness in the design and production of advanced electrical systems.
Modern warfare is defined as much by data dominance as by maneuver. From satellite-based intelligence, surveillance, and reconnaissance (ISR) platforms to dismounted soldiers' handheld radios, operational success depends on the ability to move, process, and act on digital information in real time. Yet this dependence introduces a critical vulnerability: as the force becomes more data-centric, it becomes more susceptible to disconnection, jamming, and cyber denial. In disconnected, intermittent, and limited (DIL) environments - where communications are degraded by terrain, adversarial interference, or limited infrastructure - traditional network architectures falter. Centralized command nodes and linear data pipelines cannot sustain the agility or resilience required at the tactical edge. The solution is a new design paradigm - one that integrates ruggedized hardware, edge computing, artificial intelligence (AI), and hybrid tactical-cloud architectures into a distributed, adaptive ecosystem.
In the electrical machines, detrimental effects resulted often due to the overheating, such as insulation material degradation, demagnetization of the magnet and increased Joule losses which result in decreased lifetime, and reduced efficiency of the motor. Hence, by effective cooling methods, it is vital to optimize the reliability and performance of the electric motors and to reduce the maintenance and operating costs. This study brings the analysis capability of CFD for the air-cooling of an Electric-Motor (E-Motor) powering on Deere Equipment's. With the aggressive focus on electrification in agriculture domain and based on industry needs of tackling rising global warming, there is an increasing need of CFD modeling to perform virtual simulations of the E-Motors to determine the viability of the designs and their performance capabilities. The thermal predictions are extremely vital as they have tremendous impact on the design, spacing and sizes of these motors.
As mission-critical systems demand more processing power, real-time data movement, and multi-domain interoperability, rugged embedded systems are being transformed. Today's military and aerospace applications increasingly demand the merging of AI computing, enhanced sensor interfaces, and cybersecurity - all under harsh environmental conditions. At the heart of this evolution is the 3U OpenVPX form factor, a modular, compact, and ruggedized hardware standard and increasingly the SOSA aligned subset of the architecture. However, next-generation systems need to go further: supporting higher bandwidth, better thermal efficiency, improved security, while maintaining multi-vendor interoperability and long-term sustainability. We'll discuss some of today's enclosure solutions as well as emerging technologies.
Object detection has many different uses in Command and Control (C2) systems such as autonomous control, target tracking, threat detection, and general surveillance. Graphics Processing Units (GPUs) are the de-facto standard hardware for these types of workloads in datacenter environments. Still, when deploying to an edge environment many considerations are required to ensure an optimized deployment. This paper provides a general overview of how to utilize GPUs for AI inference for object detection at the edge using NVIDIA® HoloScan as well as an overview of the many considerations to account for when selecting the most optimal GPU for any specific ground vehicle solution.
As SAE standard J3400 (also known as NACS, or North American Charging Standard) is being adopted by automakers and deployed on the latest EVs, the standard itself is still evolving. That latest evolution is SAE J3400/2. That extra 2 will make charging quicker, thanks to hardware updates to the port and inlet. As standards are announced, there are elements that result in a standard within the standard. Essentially, J3400 is more of a family of standards that handle everything from the internal technology that allows for compatibility, the hardware specs and testing of adapters that have and will be deployed and, in the case of J3400/2, the hardware itself.
This paper will present the use of a licensed open-source software application based on commercially available off-the-shelf hardware for the control and data acquisition of aerospace system integration test rigs. System integration test rigs are complex systems requiring real-time deterministic control and high-speed data acquisition. Various aircraft flight systems and subsystems can be tested to see if they interact as they would on the aircraft without an airframe. These systems are critical to ensure interoperability during the development phase and facilitate the interchangeability of actual flight hardware, prototypes, and simulation models throughout the development cycle. Deploying open, flexible, and highly configurable real-time control and data acquisition systems ensures that development milestones will be achieved cost-effectively, whether using actual flight hardware or working with a simulation. This is because, as the prototype hardware is developed, the remaining aircraft systems can still be tested by interacting with the model.
Low-level flight, defined by high-speed operations near terrain, represents a significant challenge in military rotorcraft missions while providing strategic advantages, such as radar evasion and heightened surprise. Recent conflicts highlight the urgent need for advanced low-level flight capabilities in the design of new rotorcraft. The close proximity to ground obstacles, combined with the complexities of piloting, necessitates precise control and robust handling qualities to prevent accidents. However, existing handling quality standards, such as MIL-DTL-32742, reveal limitations in assessing low-level maneuvers. Given the diverse array of new rotorcraft designs, driven by initiatives like the U.S. Army's Future Vertical Lift and NATO's Next Generation Rotorcraft Capabilities, a customized handling qualities evaluation for each design is impractical. In response, a performance-driven strategy has been implemented, scaling Mission Task Elements to align with aircraft performance capabilities. This approach identifies handling quality gaps across the Operational Flight Envelope, concentrating on the aircraft’s effectiveness in achieving task success under varied conditions. Prior simulator studies validate the effectiveness of this method for assessing different configurations. This paper presents flight test results using DLR's ACT/FHS research helicopter, confirming a set of scalable Mission Task Elements developed at DLR's AVES and NASA's VMS simulators. Pilots utilized a Head-Mounted Display for task cueing, eliminating the need for physical infrastructure. The Mission Task Elements proved suitable for evaluating the low-level handling qualities of the ACT/FHS. Although the provided Head-Mounted Display facilitated Handling Qualities evaluations, it encountered some hardware limitations. The scaling for different airspeeds met pilot expectations, and wind compensation functioned as anticipated, enhancing the independence of flight tests from environmental conditions. These findings lead to recommended updates for task descriptions and course cueing requirements, confirming desired performance tolerances.
Items per page:
50
1 – 50 of 2468