Browse Topic: Voice / speech

Items (157)
Find
A GenAI-Based Approach for Real-Time Multilingual In-Flight Communication Systems2026-26-07846/1/2026
In today’s global aviation industry, passenger experience is strongly influenced by effective communication. In-flight announcements, often limited to English and a single local language, can create confusion and stress for international travelers who may not be fluent in either. This communication gap not only impacts passenger comfort but also poses potential risks in conveying time-sensitive or safety-critical information. Recent advances in Generative Artificial Intelligence (GenAI), particularly in speech recognition, neural machine translation, and naturalistic text-to-speech, provide a pathway to overcome these challenges. This paper explores the concept of real-time multilingual in-flight announcements delivered in each passenger’s preferred language through connected headphones or personal devices. The proposed system architecture integrates speech-to-text conversion, language translation, and speech synthesis with aircraft infotainment platforms. Potential applications range from pre-generated multilingual safety messages to long-term visions of fully personalized, real-time translations with minimal latency. Benefits include improved inclusivity, accessibility for hearing-impaired passengers, and enhanced brand differentiation for airlines. Challenges such as regulatory certification, translation accuracy, latency constraints, and hardware integration must be addressed. Beyond aerospace, this capability has cross-domain relevance in automotive, railways, and public services, making it a promising area for future customer experience innovations.
Mishra, Ashwini, Kature, Kartik, Patil, Ashish
Proactive AI-Enhanced Cybersecurity for Next-Gen Automotive Infotainment ECU2026-26-06191/16/2026
With the rapid advancement of connected vehicle technologies, infotainment Electronic Control Units (ECUs) have become central to user interaction and connectivity within modern vehicles. However, this enhanced functionality has introduced new vulnerabilities to cyberattacks. This paper explores the application of Artificial Intelligence (AI) in enhancing the cybersecurity framework of infotainment ECUs. The study introduces AI-powered modules for threat detection and response, presents an integrated architecture, and validates performance through simulation using MATLAB, CANoe, and NS-3. This approach addresses real-time intrusion detection, anomaly analysis, and voice command security. Key benefits include zero-day exploit resistance, scalability, and continuous protection via OTA updates. The paper references real-world automotive cyberattack cases such as OTA vulnerability patches, Connected Drive exploits, and Uconnect hack, emphasizing the critical need for AI-enabled proactive cybersecurity frameworks.
More, Shweta, Kulkarni, Shraddha, Kumar, Priyanshu, Ghanwat, Hemant, Joshi, Vivek
Accelerating Sunroof System Development with Modular Hardware-in-the-Loop Testing2026-26-05191/16/2026
This paper presents a novel Hardware-in-the-Loop (HiL) testing framework for validating panoramic Sunroof systems independent of infotainment module availability. The increasing complexity of modern automotive features—such as rain-sensing auto-close, global closure, and voice-command operation—has rendered traditional vehicle-based validation methods inefficient, resource-intensive, and late in the development cycle. To overcome these challenges, a real-time HiL system was developed using the Real time simulation, integrated with Simulink-based models for simulation, control, and fault injection. Unlike prior approaches that depend on complete vehicle integration, this methodology enables early-stage testing of Sunroof ECU behavior across open, close, tilt, and shade operations, even under multi-source input conflicts and fault conditions. Key innovations include the emulation of real-world conditions such as simultaneous voice and manual commands, sensor faults, and environmental triggers using a software-controlled test environment. The system helps more than 60 automated test cases and makes regression testing easier without hardware reconfiguration, accelerating feedback cycles and enhancing software readiness. The results show that the framework efficiently identifies test case failures and speeds up validation timelines. The simulation model allows reuse for all ECU variants and streamlines test expansion for future functionalities. Simulation contributes a scalable and infotainment-free testing approach that enhances product quality, reduces dependency on physical prototypes, and supports continuous system integration in automotive control system.
Ghanwat, Hemant, Lad, Aniket Suryakant, Joshi, Vivek, More, Shweta
An AI-Based Embedded Offline Voice Assistant for Commercial Trucks: Enhancing Driver Safety and HMI Design2025-36-007812/18/2025
Commercial vehicle operation faces challenges from driver distraction associated with traditional Human-Machine Interfaces (HMIs) and inconsistent network connectivity, particularly in long-haul scenarios. This paper addresses these issues through the development and presentation of an embedded, offline AI-powered voice assistant. The system is designed to reduce driver distraction and enhance operational efficiency by enabling hands-free control of vehicle functions and access to critical information, irrespective of internet availability. The technical approach involves a three-tier architecture comprising an Android-based In-Vehicle Infotainment (IVI) unit for primary user interaction and voice processing, an Android mobile device acting as a communication bridge and processing hub, and a proprietary OBD-II dongle for CAN bus interfacing. Offline speech recognition is achieved using embedded wake word detection and speech-to-intent engines. A user-centered design methodology, informed by a field study with 25 professional truck drivers in Brazil, guided the prioritization of system functionalities. Key findings from this study highlighted strong driver interest in voice interaction for vehicle status monitoring (e.g., fluid levels, fault alerts) and control of essential systems (e.g., lighting, cabin environment). The implemented prototype successfully integrates these prioritized features, demonstrating the viability of offline voice control. Preliminary observations indicate robust wake word and intent recognition accuracy (≥97% based on vendor benchmarks) and acceptable system responsiveness (400-700 ms latency) under typical cabin noise conditions. This work establishes a foundation for safer, more intuitive HMIs in software-defined commercial vehicles, emphasizing the importance of offline capabilities for reliable operation.
De Oliveira Nelson, Rafael, De Almeida, Lucas Gomes, Arantes Levenhagen, Ivan
Mapping out future vehicle cockpits24AUTD05_095/1/2024
Game-like navigation visuals Conversational-style voice commands. Contactless biometric sensing. A tidal wave of software code and sensing technologies are being prepped to alter in-vehicle activities. Two supplier companies, TomTom and Mitsubishi Electric Automotive America (MEAA), recently presented their concept cockpit demonstrators to media at TomTom's North American corporate offices in Farmington Hills, Michigan. A few highlights:
Buchholz, Kami
A New U-Net Speech Enhancement Framework Based on Correlation Characteristics of Speech2024-01-20154/9/2024
As a key component of in-vehicle intelligent voice technology, speech enhancement can extract clean speech signals contaminated by environmental noise to improve the perceptual quality and intelligibility of speech. It has extensive applications in the field of intelligent car cabins. Although some end-to-end speech enhancement methods based on time domain have been proposed, there is often limited consideration given to designing model architectures based on the characteristics of the speech signal. In this paper, we propose a new U-Net based speech enhancement framework that utilizes the temporal correlation of speech signals to reconstruct higher-quality and more intelligible clean speech. Firstly, to address the issue of inadequate extraction of multi-scale correlation features from speech signals during feature extraction and reconstruction, a novel dense connection multi-scale feature extraction module based on gated dilated convolution is devised to enhance temporal receptive length and extract diverse scale features effectively. Secondly, in order to tackle the problem of feature loss and harmonic distortion during sampling, a sophisticated pooling-reconstruction fine-grained sampling method based on feature map recombination is proposed. This method aims to minimize information loss during down-sampling while simultaneously enhancing the clarity of reconstructed waveforms during up-sampling. Lastly, leveraging the aforementioned pooling-reconstruction sampling method, we propose a deep supervision approach for multi-scale feature. This approach effective supervision of perception characteristics across different frequency ranges. In order to validate the effectiveness of the proposed framework, experiments were conducted on the Voicebank+Demand dataset. The results show that compared to other advanced algorithms, the proposed model significantly improves metrics such as PESQ, STOI, CSIG, CBAK, and COVL. Even in low SNR environments, the enhanced speech signals exhibit noticeable improvements in quality and intelligibility. This is beneficial for subsequent automotive voice applications.
Zhang, Lijun, Pei, Kaikun, Li, Wenbo, Meng, Dejian, He, Yinzhi
A Target-Speech-Feature-Aware Module for U-Net Based Speech Enhancement2024-01-20214/9/2024
Speech enhancement can extract clean speech from noise interference, enhancing its perceptual quality and intelligibility. This technology has significant applications in in-car intelligent voice interaction. However, the complex noise environment inside the vehicle, especially the human voice interference is very prominent, which brings great challenges to the vehicle speech interaction system. In this paper, we propose a speech enhancement method based on target speech features, which can better extract clean speech and improve the perceptual quality and intelligibility of enhanced speech in the environment of human noise interference. To this end, we propose a design method for the middle layer of the U-Net architecture based on Long Short-Term Memory (LSTM), which can automatically extract the target speech features that are highly distinguishable from the noise signal and human voice interference features in noisy speech, and realize the targeted extraction of clean speech. Then, in order to achieve deep fusion between the target speech features and the model, we design a multi-scale deep fusion skip connection method, so that when the effective information flows from the encoder to the decoder, the features with large correlation with the target speech are effectively screened through the weight coefficient of attention. Finally, in order to verify the effectiveness of the proposed module, experiments were carried out on the Voicebank+Demand speech dataset. The results show that the proposed method has strong robustness in the environment with human voice interference. It outperforms other algorithms on metrics such as PESQ, STOI, CSIG, CBAK, COVL, offering cleaner speech with higher perceptual quality and intelligibility. This makes it particularly promising for applications in scenarios with significant human voice interference, such as in-car environments.
Pei, Kaikun, Zhang, Lijun, Meng, Dejian, He, Yinzhi
CES 2024: VW, Cerence integrate ChatGPT in IDA voice assistant24AUTP02_082/1/2024
ChatGPT has entered the car. At CES 2024, Volkswagen and technology partner Cerence introduced an update to IDA, VW's in-car voice assistant, so it can now use ChatGPT to expand what's possible using voice commands in vehicles. VW said the ChatGPT bot will be available in Europe in current MEB and MQB evo models from VW Group brands that currently use the IDA voice assistant. That includes some members of the ID family - the ID.7, ID.4, ID.5 and ID.3 - as well as the new Tiguan, Passat and Golf models. VW brands Seat, Škoda, Cupra and VW Commercial Vehicles also will get IDA integration. VW hopes to bring IDA to other markets, including North America, but did not make any timing announcements.
Blanco, Sebastian
The Utilization of Psychometric Functions to Predict Speech Intelligibility in Vehicles10-08-01-000212/29/2023
In this study, a novel assessment approach of in-vehicle speech intelligibility is presented using psychometric curves. Speech recognition performance scores were modeled at an individual listener level for a set of speech recognition data previously collected under a variety of in-vehicle listening scenarios. The model coupled an objective metric of binaural speech intelligibility (i.e., the acoustic factors) with a psychometric curve indicating the listener’s speech recognition efficiency (i.e., the listener factors). In separate analyses, two objective metrics were used with one designed to capture spatial release from masking and the other designed to capture binaural loudness. The proposed approach is in contrast to the traditional approach of relying on the speech recognition threshold, the speech level at 50% recognition performance averaged across listeners, as the metric for in-vehicle speech intelligibility. Results from the presented analyses suggest the importance of considering speech recognition accuracy across a range of signal-to-noise ratios rather than the speech recognition threshold alone, and the importance of considering individual differences among listeners when evaluating in-vehicle speech intelligibility.
Samardzic, Nikolina, Lavandier, Mathieu, Shen, Yi
EDITORIAL: Of moose and artificial intelligence23AUTP09_049/1/2023
I know nothing more about artificial intelligence (AI) than what I read and what learned people tell me. I know it's supposed to bring new sophistication to all manner of processes and technologies, including automated driving. So, when a driverless robotaxi operated by GM's Cruise plowed into a road section of freshly poured cement in San Francisco, it raised questions about recently beleaguered Cruise. My mind wandered to AI, which many AV compute “stacks” are touted to leverage in abundance. Driving into wet cement isn't intelligent. Did somebody need to train the vehicle's AV stack specifically to recognize wet cement? If that's how it works, I'd prefer not to bet my life on whether some fairly oddball happenstance (is the term ‘edge case’ not cool anymore?) had been accounted for in that particular version of the AD system's algorithm running that particular day.
Visnic, Bill
LiDAR-Based Fail-Safe Emergency Maneuver for Autonomous Vehicles2023-01-05784/11/2023
Although SAE level 5 autonomous vehicles are not yet commercially available, they will need to be the most intelligent, secure, and safe autonomous vehicles with the highest level of automation. The vehicle will be able to drive itself in all lighting and weather conditions, at all times of the day, on all types of roads and in any traffic scenario. The human intervention in level 5 vehicles will be limited to passenger voice commands, which means level 5 autonomous vehicles need to be safe and capable of recovering fail operational with no intervention from the driver to guarantee the maximum safety for the passengers. In this paper a LiDAR-based fail-safe emergency maneuver system is proposed to be implemented in the level 5 autonomous vehicle. This system is composed of an external redundant 3600 spinning LiDAR sensor and a redundant ECU that is running a single task to steer and fully stop the vehicle in emergency situations (e.g., vehicle crash, system failure, sensor failures, vehicle pile-up, etc.) by creating a map and occupancy grid of the LiDAR PointClouds, plan a path and follow the path to full stop in a short time and safe manner.
Alrousan, Qusay, Alzu'bi, Hamzeh, Tasky, Tom, Varasquim, Juliano
Understanding User Requirements for Smart Cockpit of New Energy Vehicles: A Natural Language Process Approach2022-01-707512/22/2022
The smart cockpit has become an irreplaceable element for many new automobile brands, particularly New Energy Vehicles (NEV) of “new forces”. Since the cockpit is a direct interface for the interactions between users and the intelligent and connected functions of the vehicle, any improvements would be easily perceived by users and thus would directly affect user experiences. It would be most important to capture, collect, and understand what users need for a smart cockpit. Users’ online comments on existing smart cockpits contain information on users’ requirements. However, the current user comment text data is too massive, tanglesome, and sparse to process. How to efficiently mine valuable information from these data is non-trivial. This paper focuses on applying the Natural Language Process (NLP) technology for design, development, improvement, and update of a vehicle company’s smart cockpit. By obtaining user comment data from various sources such as eco-system Applications (APP), forums, posts, Questions and Answers(Q&A), customer services, etc., we aim to mine and quantify user demand for the smart cockpit. A deep learning NLP model named Bidirectional Encoder Representations from Transformers (BERT) is developed. In addition, the incremental pre-trained BERT is proposed to predict the mentioned cockpit feature, user’s intention, and emotion from a comment, with one year’s data from our cooperated NEV company for model training. Experiment results showed that our model outperforms the conventional BERT in terms of predictive ability and consumed time. Applications in the cooperated company were discussed.
Lin, Shenhe, Zou, Jingkai, Zhang, Chaokai, Lai, Xinjun, Mao, Ning, Fu, Hui
Implementation of a Voice-Assisted Collaborative Robot in a Production Assembly Line2022-28-058911/18/2022
As the world is moving from a manual workforce to a robot-based workforce, there is a huge scope for improved methods to make production lines more efficient. In this work, an effort is made to implement human-robot collaboration into an industrial process and is demonstrated with a flange assembly-line model. This paper explains how the Yolov4 algorithm was improved and fine-tuned to meet the requirements. A customized workspace was designed and manufactured to make the components more accessible. Different types of grippers were compared and the simplest and most efficient was then selected. Camera selection and calibration were done to get the RGB coordinates and the depth values which were finally converted into the robot's coordinate frame. The coordinates are then fed as the end goal position for the end effector to which the robot plans its motion and then executes. The paper also explains how the model responds to voice commands using the Google API to convert audio messages to text hence making it easier for the operator to issue commands to the robot. The merits and demerits of the implemented system were also noted.
Seby, Harrison, Saju George, Albin, Sadique, Anwar, P P, Lalu
The Digital Cockpit of the FutureTBMG-463558/1/2022
Dynamically Managing Task Allocation Between Humans and Machines in Surveillance Operations22AERP06_096/1/2022
Constructing an Autonomous Manager (AM) for use as an integral component of distributing multiple tasks between humans and autonomous agents, particularly in Intelligence, Surveillance, and Reconnaissance (ISR) applications. Air Force Research Laboratory, Wright-Patterson Air Force Base, Ohio Increasingly sophisticated technology must be leveraged in surveillance environments to enable eventually achieving the goal of allowing analysts to increase throughput by managing multiple simultaneous feeds. Maintaining this increased tasking will likely introduce additional workload and fatigue. Fortunately, analysts can currently offload some of these tasks to automation and will, in the future, be able to offload additional tasking to streamline the intelligence analysis process. Currently, various speech-to-text and text-to-speech programs can be used to convert spoken information into chat and automation can be used to copy text to multiple needed locations simultaneously. Automation has aided in the transmission of information between analysts and organizations. Tools are also being developed to augment the detection of important visual features within surveillance scenes. However, the degree of assistance autonomous systems can provide is still somewhat limited for cognitively complex tasks, but progress is being made incrementally toward viable assistive tools. Balancing analyst workload while maintaining multiple tasks will require intelligent and dynamic distribution of tasks between humans and autonomy.
Multi-part Analysis and Techniques for Air Traffic Speech Recognition01-15-02-00145/25/2022
The general English speech recognition is based on the techniques of n-grams where the words before and after are predicted and the utterance prediction is produced. At the same time, having a significantly lengthier n-gram has its own impact in training and the accuracy. Shorter n-grams require the utterances to be split and predicted than using the complete utterance. This article discusses specific techniques to address the specific problems in Air Traffic Speech, which is a medium length utterance domain. Moving from the adapted language models (LMs) to rescored LM, a combined technique of syntax analysis along with a deep learning model is proposed, which improves the overall accuracy. It is explained that this technique can help to adapt the proposed method for different contexts within the same domain and can be successful.
Srinivasan, Narayanan, Balasundaram, S. R.
Robot based Automated Testing of Automotive InfotainmentSAE-PP-0021212/19/2021
As Automotive Infotainment device increase in feature complexity, validation becomes equally difficult. This paper describes an robot-based method to automatically test the device along the development cycle, with easy to write test cases using keyword-based language. The solution proposed is an end-to-end automation from test management, test case writing, automated validation and test reporting. This paper describes the method adopted along with key results and advantages.
Kalidas, Sateesh
Letter from the Special Issue Editors12-04-03-00178/19/2021
Letter from the Special Issue Editors
Singh, Kanwar Bharat, Arat, Mustafa Ali, Unterreiner, Michael
Intelligent Voice Activated Drone(s) for in-Vehicle Services and Real-Time Predictions2021-01-00634/6/2021
Today, commercially available drones have limited use-cases in the rapidly evolving community. However, with advances in drone and software technology, it is possible to utilize these aerial machines to solve problems in a variety of industries such as mining, medical, construction, and law enforcement. For example, in order to reduce time of investigation, Indiana State Police are currently utilizing ad-hoc commercial drones to reconstruct crash scenes for insurance and legal purposes. In this paper, we illustrate how to effectively integrate drones for in-vehicle services and real-time prediction for automotive applications. In order to accomplish this, we first integrate simpler controls such as voice-commands to control the drone from the vehicle. Next, we build smart prediction software that monitors vehicle behavior and reacts in real-time to collisions. Furthermore, we employ object recognition techniques through In-Vehicle Infotainment (IVI) systems to identify the surroundings based on inputs from drone-mounted camera sensors. Consequently, we implement object identification and smart maneuver of the drone in relation to the vehicle; as well, employ timely deployment of the drone prior to collision for emergency assistance and crash reconstruction purposes. The goal is to optimize performance and amplify safety and security of the vehicle. The prototype detailed in this paper was tested on a vehicle moving at a speed of 45 mph. The driver of the vehicle can deploy and control the drone using voice commands. The drone follows the vehicle and is in-sync with the vehicle and performs tasks to aid in post-collision assistance and crash reconstruction.
Nithiyanantham, Mayunthan, Sinnapolu, Giribabu
Electronic Display of Aeronautical Information (Charts)ARP5621A (Current)12/9/2020
The objective of this ARP is to provide a set of user-centered design guidelines for the implementation of data driven electronic aeronautical charts, which dynamically create charts from a database of individual elements. The data driven chart is intended to provide information required to navigate, but it is not intended to supplant the aircraft’s primary navigation display. These guidelines seek to provide a balance between standardization of equipment with similar intended functions and individual manufacturer innovation. This ARP provides guidelines for the display of an electronic chart that can replace existing paper. This document addresses what information is required, when it is required, and how it should be displayed and controlled. This document does not include all the detailed specifications required to generate an electronic aeronautical chart. This document primarily addresses the human factors aspects of electronic chart display, and does not address the software, hardware or system integrity/availability issues associated with certification of an electronic chart system. During the transition to data driven charts, the guidelines of this document should be applied to interim electronic chart products that may be pre-composed, such as vector or raster based electronic charts. This document is designed primarily for IFR Aeronautical Charts. There is a limited discussion of its applicability to VFR charts.
G-10EAB Executive Advisory Group
TOWARD A SOFTWARE BASED INTERCOMMUNICATIONS CAPABILITY FOR ARMY GROUND VEHICLES2024-01-37598/13/2019
ABSTRACT The confluence of intra-vehicle networks, Vehicular Integration for (C4ISR) Command, Control Communication, Computers, Intelligence, Surveillance, Reconnaissance/(EW) Electronic Warfare Interoperability (VICTORY) standards and onboard general-purpose processors creates an opportunity to implement Army combat ground vehicle intercommunications (intercom) capability in software. The benefits of such an implementation include 1) SWAP savings, 2) cost savings, 3) simplified path to future upgrades and 4) enabling of potential new capabilities such as voice activated mission command. The VICTORY Standards Support Office (VSSO), working at the direction of its Executive Steering Group (ESG) members (Program Executive Office (PEO) Ground Combat Systems (GCS), PEO Combat Support and Combat Service Support (CS&CSS), PEO Command Control Communications-Tactical (C3T) and PEO Intelligence, Electronic Warfare and Sensors (IEW&S)), has developed and demonstrated a software intercom prototype that proves out the concept and sets the stage for development of a deployable software intercom capability. This paper describes that effort to date including benefits to the Army, technical trades explored and potential for extended capabilities.
Kelsch, Geoffrey, Serafinko, Robert, Frissora, Anthony
Automatic Speech Recognition System Considerations for the Autonomous Vehicle2019-01-08614/2/2019
As automakers begin to design the autonomous vehicle (AV) for the first time, they must reconsider customer interaction with the Automatic Speech Recognition (ASR) system carried over from the traditional vehicle. Within an AV, the voice-to-ASR system needs to be capable of serving a customer located in any seat of the car. These shifts in focus require changes to the microphone selection and placement to serve the entire vehicle. Further complicating the scenario are new sources of noise that are specific to the AV that enable autonomous operation. Hardware mounted on the roof that are used to support cameras and LIDAR sensors, and mechanisms meant to keep that hardware clean and functioning, add even further noise contamination that can pollute the voice interaction. In this paper, we discuss the ramifications of picking up the intended customer’s voice when they are no longer bound to the traditional front left “driver’s” seat. Considerations are made to the possibilities of new microphone construction and layouts to provide coverage for all potential passengers, and cost-efficient minimal microphone packages are discussed. Additionally, if the automaker chooses to initiate the ASR interaction with a “wake up word”, instead of installing Push to Talk (PTT) buttons for every seat, we discuss how the multiple microphone’s placements can be leveraged to identify the seat issuing the command, and focus further ASR interactions with that location in the car.
Wheeler, Joshua
Analysis of Automatic Speech Recognition Failures in the Car2019-01-03974/2/2019
In this paper, an approach to analyze voice recognition data to understand how customers use voice recognition systems is explored. The analysis will help identify ASR failures and usability related issues that customers encounter while using the voice recognition system. This paper also examines the impact of these failures on the individual speech domains (media control, phone, navigation, etc.). Such information can be used to improve the current voice recognition system and direct the design of future systems. Infotainment system logs, audio recordings of the voice interactions, their transcriptions and CAN bus data were identified to be rich sources of data to analyze voice recognition usage. Infotainment logs help understand how the system interpreted or responded to customer commands and at what confidence level. The audio recordings of the voice interaction and their transcriptions provide information about what command is issued by the customer and if it adheres to the grammar of the voice recognition system. The system’s interpretation of the command from the logs can be compared to the actual command issued to detect if it is correctly recognized by the system. CAN bus data can help in determining if voice recognition failures occur due to noise sources in the car such as HVAC blower noise, engine noise, etc. These data sources can also be tied together to detect commands that are incorrectly recognized by the system. When the causes of failures by domain were studied, it was found that the navigation domain was most prone to errors. Natural language understanding and single command navigation would improve the success of the navigation domain. The media control and phone domains were significantly less error-prone. Errors that occurred were largely due to core speech recognition and a majority of those errors could be handled by examining the participant’s habits.
Rangarajan, Ranjani, Amman, Scott, Busch, Leah
Aero-Vibro-Acoustic Simulation Methodologies for Vehicle Wind Noise Reduction2019-26-02021/9/2019
Wind noise is a major contributor to vehicle noise and a very common consumer complaint for overall vehicle quality [1]. The reduction of wind noise is becoming even more important as powertrain noise is reduced or eliminated (by conversion to hybrid and electric vehicles) and as the importance of quiet interior environment for hands-free device use and voice activation systems becomes more pronounced. In contrast to other noise sources such as tires, engine, intake, exhaust or other component noise whose acoustic loads may be measured in a direct and standardized way with the proper equipment, wind noise is very difficult to predict because the acoustic part of wind noise is a small component of overall fluctuating pressures. It is very challenging to either directly measure or to simulate the acoustic component of fluctuating exterior pressures using CFD (Computational Fluid Dynamics) without a great deal of specialized experience in this area. This paper addresses the challenges of Aero-Vibro-Acoustics (AVA) modeling specifically for vehicle wind noise applications and describes what types of CFD analyses are suitable for driving vibro-acoustic models to predict wind noise and what methods for converting wind tunnel test data or CFD outputs to acoustic input loads are most effective. The various vibro-acoustic modeling techniques that have been successfully employed to predict and optimize interior noise via sensitivity studies of interior sound package and glass damping treatments are described. Validation examples and comparisons are shown and conclusions about current best modeling practices and future areas of investigation are presented:
Musser, Chadwyck, Calloni, Massimiliano, Golota, Anton, Zerbib, Nicolas
WHAT WE'RE DRIVING18AUTP10_0510/1/2018
Mercedes-Benz is fond of saying it cascades at least some of the S-Class' posh luxury and technology throughout its range. That's just what it has done for the fourth-generation A-Class, the nameplate that sits at the bottom of the brand's lineup. The new A-Class is heaped with cabin tech, in particular. Built on Mercedes' MFA2 front-wheel-drive platform, the A200 AMG Line version driven is well-connected, having the MBUX (Mercedes-Benz User Experience) touchscreen interface (and a touchpad), plus “Hey Mercedes” voice recognition.
Educating UGVsTBMG-289105/1/2018
Advancements in the field of artificial intelligence (AI) are accelerating as the technology matures from being research-orientated to being deployed in a wide range of products and services, such as autonomous vehicles. While Convolutional Neural Networks (CNNs) were first described in the 1950s, the technology remained an academic concept until the availability of large training data sets and powerful Graphics Processor Units (GPUs), a processor architecture ideal for the heavy math computational demands associated with neural network processing. Once scientists had low-cost and high-performance platforms, the technology exploded for many commercial uses. Military use is more challenging due to the lack of large data sets, but that is changing too as areas including thermal imagery are starting to be used.
Sound Inaudible to Humans Can Be Recorded by MicrophonesTBMG-287024/1/2018
Microphones, from those in smartphones to hearing aids, are built specifically to hear the human voice — humans can’t hear at levels higher than 20 kHz, and microphones max out at around 24 kHz, meaning that microphones only capture the sound humans hear.
Analysis of Vehicle Voice Recognition Performance in Response to Background Noise and Gender Based Frequency2017-01-18886/5/2017
Voice Recognition (VR) systems have become an integral part of the infotainment systems in the current automotive industry. However, its recognition rate is impacted by external factors such as vehicle cabin noise, road noise, and internal factors which are a function of the voice engine in the system itself. This paper analyzes the VR performance under the effect of two external factors, vehicle cabin noise and the speakers’ speech patterns based on gender. It also compares performance of mid-level sedans from different manufacturers.
Khan, Rasheed, Ali, Mahdi, Frank, Eric C.
The Effect of Vehicle Noise on Automatic Speech Recognition Systems2017-01-18646/5/2017
The performance of a vehicle’s Automatic Speech Recognition (ASR) system is dependent on the signal to noise ratio (SNR) in the cabin at the time a user voices their command. HVAC noise and environmental noise in particular (like road and wind noise), provide high amplitudes of broadband frequency content that lower the SNR within the vehicle cabin, and work to mask the user’s speech. Managing this noise is a vital key to building a vehicle that meets the customer’s expectations for ASR performance. However, a speech recognition engineer is not likely to be the same person responsible for designing the tires, suspension, air ducts and vents, sound package and exterior body shape that define the amount of noise present in the cabin. If objective relationships are drawn between the vehicle level performance of the ASR system, and the vehicle or system level performance of the individual noise, vibration and harshness (NVH) attributes, a partnership between the groups is brokered. Compatible targets are set and hardware selected that works to meet both groups’ goals. This paper examines the NVH attributes and performance metrics that relate to vehicle level ASR performance, and finds that strong relationships and statistical trends can be drawn between the Sentence Error Rate (SER%) and standard NVH metrics for that road surface or HVAC configuration. The paper also establishes that AI% should be the preferred metric to relate cabin noise to ASR performance in the presence of any other kind of steady state noise.
Wheeler, Joshua
The Effect of HVAC Buffeting on Automatic Speech Recognition Systems2017-01-17816/5/2017
The design and operation of a vehicle’s heating, ventilation, and air conditioning (HVAC) system has great impact on the performance of the vehicle’s Automatic Speech Recognition (ASR) and Hands-Free Communication (HFC) system. HVAC noise provides high amplitudes of broadband frequency content that affects the signal to noise ratio (SNR) within the vehicle cabin, and works to mask the user’s speech. But what’s less obvious is that when the airflow from the panel vents or defroster openings can be directed toward the vehicle microphone, a mechanical “buffeting” phenomenon occurs on the microphone’s diaphragm that distresses the ASR system beyond its ability to interpret the user’s voice. The airflow velocity can be strong enough that a simple windscreen on the microphone is not enough to eliminate the problem. Minimizing this buffeting effect is a vital key to building a vehicle that meets the customer’s expectations for ASR and HFC performance. Systems design principles must be applied to ensure that the placement of the microphone and vents, HVAC airflow management, and active noise reduction solutions are all designed in concert to reduce exposure to the problem. In this paper, we examine the objective effect that HVAC buffeting has on the ASR system, above and beyond the masking noise provided when the airflow is directed away from the microphone. We discuss vent and microphone placement that can contribute to the error state, and propose design guidelines or active solutions that can help reduce the occurrence and impact of HVAC buffeting.
Wheeler, Joshua
Low-Power, Special-Purpose Chip for Speech Recognition in ElectronicsTBMG-268635/1/2017
Automatic speech recognition is on the verge of becoming the chief means of interacting with computing devices. To address this, MIT researchers have built a low-power chip specialized for automatic speech recognition. Whereas a cellphone running speech recognition software might require about 1 Watt of power, the new chip requires between 0.2 and 10 milliwatts, depending on the number of words it has to recognize. That probably translates to a power savings of 90 to 99 percent, which could make voice control practical for relatively simple electronic devices, including power-constrained devices that harvest energy from their environments, or go months between battery charges.
Developing a Multi-Modal UGV Robot Control Interface17AERP05_045/1/2017
Unmanned ground vehicles (UGV) are currently being used or developed for operational maneuvers (e.g., reconnaissance and IED defeat), maneuver support (e.g., route clearance), and sustainment (e.g. convoy and resupply) missions. Despite the demonstrable benefits of UGVs, significant challenges remain to their effective integration into military operations. Currently-fielded UGVs require active remote control or teleoperation, even for mundane tasks such as accompanying a soldier or vehicle during maneuvers. While UGVs have proven invaluable in removing warfighters from harm's way, most notably in explosive ordnance disposal missions, the current reliance on teleoperation strongly restricts the number of platforms that can be deployed and the range of missions that can be effectively supported.
Developing a Multi-Modal UGV Robot Control InterfaceTBMG-268285/1/2017
Unmanned ground vehicles (UGV) are currently being used or developed for operational maneuvers (e.g., reconnaissance and IED defeat), maneuver support (e.g., route clearance), and sustainment (e.g. convoy and resupply) missions. Despite the demonstrable benefits of UGVs, significant challenges remain to their effective integration into military operations.
Validation of In-Vehicle Speech Recognition Using Synthetic Mixing2017-01-16933/28/2017
This paper describes a method to validate in-vehicle speech recognition by combining synthetically mixed speech and noise samples with batch speech recognition. Vehicle cabin noises are prerecorded along with the impulse response from the driver's mouth location to the cabin microphone location. These signals are combined with a catalog of speech utterances to generate a noisy speech corpus. Several factors were examined to measure their relative importance on speech recognition robustness. These include road surface and vehicle speed, climate control blower noise, and driver's seat position. A summary of the main effects from these experiments are provided with the most significant factors coming from climate control noise. Additionally, a Signal to Noise Ratio (SNR) experiment was conducted highlighting the inverse relationship with speech recognition performance.
Huber, John, Rangarajan, Ranjani, Ji, An, Charette, Francois, Amman, Scott, Wheeler, Joshua, Richardson, Brigitte
The Impact of Microphone Location and Beamforming on In-Vehicle Speech Recognition2017-01-16923/28/2017
This paper describes two case studies in which multiple microphone processing (beamforming) and microphone location were evaluated to determine their impact on improving embedded automatic speech recognition (ASR) in a vehicle hands-free environment. While each of these case studies was performed using slightly different evaluation set-ups, some specific and general conclusions can be drawn to help guide engineers in selecting the proper microphone location and configuration in a vehicle for the improvement of ASR. There were some outcomes that were common to both dual microphone solutions. When considering both solutions, neither was equally effective across all background noise sources. Both systems appear to be far more effective for noise conditions in which higher frequency energy is present, such as that due to high levels of wind noise and/or HVAC (heating, ventilation and air conditioning) blower noise. Microphone location was also shown to have a substantial effect on the performance of the ASR system. The results from both studies showed that simply moving a single microphone from the overhead console (OHC) to the sun visor near the driver can provide a great benefit in ASR performance without the cost of multiple beamforming microphones. For the sole purposes of speech enhancement of the driver, it is recommended that moving a single microphone from the OHC to the driver’s sun visor position be performed before additional microphones are added. However, beamforming could provide a role if the desire is to include other occupants in a voice session. For instance, a centrally located beamformer could be steered toward the vehicle occupant wishing to issue a voice command. Ultimately, a user-case strategy of the voice control application will determine the microphone configuration and location(s).
Amman, Scott, Huber, John, Charette, Francois, richardson, Brigitte, Wheeler, Joshua
Systems Engineering Approach for Voice Recognition in the Car2017-01-15993/28/2017
In this paper, a systems engineering approach is explored to evaluate the effect of design parameters that contribute to the performance of the embedded Automatic Speech Recognition (ASR) engine in a vehicle. This includes vehicle designs that influence the presence of environmental and HVAC noise, microphone placement strategy, seat position, and cabin material and geometry. Interactions can be analyzed between these factors and dominant influencers identified. Relationships can then be established between ASR engine performance and attribute performance metrics that quantify the link between the two. This helps aid proper target setting and hardware selection to meet the customer satisfaction goals for both teams.
Wheeler, Joshua, Richardson, Brigitte, Amman, Scott, Ji, An, Huber, John, Rangarajan, Ranjani
Computer Chips Calculate and Store in an Integrated UnitTBMG-265183/1/2017
Researchers at the University of Wisconsin-Madison created computer chips that can be configured to perform complex calculations and store massive amounts of information within the same integrated unit, and communicate efficiently with other chips. Called “liquid silicon” — liquid for software and silicon for hardware — the technology has uses in data-intensive applications such as facial or voice recognition, natural language processing, and graph analytics.
CONSUMER ELECTRONICS COME ON BOARD17TOFHP02_032/1/2017
Smartphones and tablets are likely to play a large role in HMIs for heavy vehicles-if productivity and safety are not compromised. The tablets and smartphones that transformed consumer electronics are now poised to enter heavy-duty vehicle markets. They're already becoming part of the maintenance and diagnostic world, while some developers are exploring ways to integrate them into human-machine interfaces (HMIs) for non-safety activities. It's almost a given that HMIs for off-highway vehicles and trucks will move closer to the management techniques used in tablets and smartphones. The question will be how to get the most from these systems without reducing productivity and safety. Developers are creating strategies and products that borrow from these devices and include them.
Costlow, Terry
Electronic Device Monitors the Heart and Recognizes SpeechTBMG-263352/1/2017
Researchers from the University of Colorado Boulder and Northwestern University have developed a tiny, soft, and wearable acoustic sensor that measures vibrations in the human body, allowing them to monitor human heart health and recognize spoken words. The stretchable device captures physiological sound signals from the body, has physical properties matched with human skin, and can be mounted on nearly any surface of the body. The sensor resembles a small Band-Aid®, weighs less than one-hundredth of an ounce, and can gather continuous physiological data.
SAE Convergence 201616AUTP11_0111/1/2016
Talk of the healthy aspects of disruption mingles with SAE's renowned technical emphasis to foster the auto industry's continuing evolution toward electrification and autonomy. The auto industry is at the moment ravenous for discussion and guidance regarding the ongoing integration of electronics into vehicle development-but particularly regarding the almost daily deluge of news and speculation about vehicle automation. SAE's 2016 Convergence conference supplied a diverse forum for those conversations while providing a new format that grouped each of the conference's three days into a distinct subject-matter theme. From an insightful, consumer-focus keynote address from Ford's Raj Nair, Executive Vice President, Product Development and Chief Technical Officer, to technical sessions that explored everything from trends in human-machine interface to supply-chain transformation, Convergence once again was the rallying point for technical and strategic discussion at a high level.
Costlow, Terry, Visnic, Bill
Reducing Sensor and Readout Circuitry Noise in Digital Domain Using Reference PixelsTBMG-2552210/1/2016
One of the main heritage tools used in scientific and engineering data spectrum analysis is the Fourier Integral Transform and its high-performance digital equivalent — the Fast Fourier Transform (FFT). The Fourier view of nonlinear mechanics that has existed for a long time and the associated FFT carry strong a-priori assumptions about the source data, such as linearity and being stationary. Natural phenomena measurements are essentially nonlinear and nonstationary.
Interactive Modes F-ANP Evaluation for In-Vehicle Secondary Tasks2016-01-18909/14/2016
With the development of automotive HMI and mobile internet, many interactive modes are available for drivers to fulfill the in-vehicle secondary tasks, e.g. dialing, volume adjustment, music playing. For driving safety and drivers’ high expectation for HMI, it is urgent to effectively evaluate interactive mode with good efficiency, safety and good user experience for each secondary tasks, e.g. steering wheel buttons, voice control. This study uses a static driving simulation cockpit to provide driving environment, and sets up a high-fidelity driving cockpit based on OKTAL SacnerStudio and three-dimensional modeling technology. The secondary tasks supported by HMI platform are designed by customer demands research. The secondary task test is carried out based on usability test theory, and the influence on driving safety by different interactive modes is analyzed. By F-ANP fuzzy network analysis method, the different influence factors of secondary task interactive modes are taken into consideration comprehensively, and an evaluation model is set up, while safety and usability are the first level index. Then a comprehensive evaluation is carried out on different secondary tasks, and the model is verified according to customer satisfaction degree. Finally, the evaluation result is tentatively integrated with the user population characteristics to do the correlation analysis. It is found that the score of in-vehicle secondary task interaction experience by the same interaction method has certain relationship with user gender and age. This paper offers reference for the design, application and evaluation of different new HMI technologies in the car.
Ma, Jun, Gong, Zaiyan, Dong, Yiwei
All-new 2017 Porsche Panamera moves to VW modular platform16AUTP09_149/1/2016
Creating any new Porsche is one of the most challenging design and engineering tasks in the auto industry. Whether sports car, sedan or SUV it has to embody the essence of the iconic 911 but also meet the practical and aesthetic expectations of buyers who believe they know exactly what they want. In the view of some critics, the 2009 Panamera sports sedan-or 4-door Gran Turismo as Porsche describes it-did not wholly meet those criteria, although with more than 150,000 built it was hardly a failure.
Birch, Stuart
Harman advanced concept brings intuitive HMI to automated driving16MEIP06_126/1/2016
Harman International is demonstrating an advanced connected-car platform that previews a fundamental shift in the human-machine interface of automated and autonomous vehicles. For this project Harman is collaborating with Swiss-based concept vehicle specialist Rinspeed, which developed the BMW i8-based Etos demonstrator featuring V2X (vehicle to everything) capability and incorporating the latest Microsoft driver “assistant.” The Harman platform, known as LIVS (Life-Enhancing Intelligent Vehicle Solution), is based on RoadLINK technology from NXP Semiconductors. The demonstrator shows how vehicle occupants will safely interact with the outside world, whether driving in SAE Level 2 or 3 automated mode, or in fully autonomous SAE Level 5.
Adcock, Ian
CARS POISED TO BECOME ‘A THING’16MEIP06_036/1/2016
Making automobiles part of the Internet of Things brings both risks and rewards. The Internet of Things (IoT) will be the next transformational aspect of connectivity, enabling machines to communicate with each other without human intervention. Autos are becoming a major player in the IoT era, talking to cloud computers and potentially to other vehicles. IoT connectivity requires automakers to adopt common standards and rely on independent service suppliers. This openness brings significant benefits-but also risks.
Costlow, Terry
CARS POISED TO BECOME ‘A THING’16AUTP03_023/1/2016
Making ‘A THING’ automobiles part of the Internet of Things brings both risks and rewards. The Internet of Things (IoT) will be the next transformational aspect of connectivity, enabling machines to communicate with each other without human intervention. Autos are becoming a major player in the IoT era, talking to cloud computers and potentially to other vehicles. IoT connectivity requires automakers to adopt common standards and rely on independent service suppliers. This openness brings significant benefits-but also risks.
Costlow, Terry
Sub-Audible Speech Recognition Based on Electromyographic (EMG) SignalsTBMG-239082/1/2016
Sub-audible speech is a new form of human communication that uses tiny neural impulses (EMG signals) in the human vocal tract instead of audible sounds. These EMG signals arise from commands sent by the brain’s speech center to tongue and larynx muscles that enable production of audible sounds. Sub-audible speech arises from EMG signals intercepted before an audible sound is produced and, in many instances, allows inference of the corresponding word or sound. Where sub-audible speech is received and appropriately processed, production of recognizable sounds is no longer important. Further, the presence of noise and of intelligibility barriers, such as accents associated with the audible speech, no longer hinder communication.
TIME TO GET personal16AUTD01_011/1/2016
Industry engineers are combining apps, voice, the cloud, and other technologies such as artificial intelligence to enable drivers to customize their vehicles and anticipate their needs. The race to let drivers personalize vehicles is gaining speed. A range of techniques and technologies are being deployed to let drivers do more to make vehicles as individualized as their consumer environment. Design teams are pushing personalization well beyond simple tasks like moving seats and picking radio stations to match instructions in key fobs. Apps and other techniques that personalize rental and ride-sharing vehicles are emerging, as are technologies that alter settings even when multiple people share a key fob.
Costlow, Terry
SAE Eye on Engineering: Voice-Activated Commands Study by AAA1207311/3/2015
Voice-activated commands have been engineered into new vehicles to help drivers keep their eyes on the road and reduce driver distraction. But a recent study by AAA's Foundation for Traffic Safety has found that voice command technology is not as effective as hoped. In this episode of SAE Eye on Engineering, Senior Editor Lindsay Brooke takes a look at the findings of AAA's study. SAE Eye on Engineering also airs Monday mornings on WJR 760 AM Detroit's Paul W. Smith Show. Access archived episodes of SAE Eye on Engineering at: http://www.sae.org/magazines/podcasts.
Audi chooses high technology, cautious design evolution for new A415AUTP10_0510/1/2015
In addition to lighter weight and significant improvements in efficiency, the new car employs plenty of technology and driver support. Audi's just-revealed 2016MY A4 is a classic example of automotive evolutionary design. Its aesthetics are not quite ripe for a game of “spot the difference” between the new fifth and previous fourth generation of this hugely successful model line, but they are close to it, even though every panel has been changed. What this car is about, is evolution, technology, and systems' development; development of practically every aspect of its being (including new engine technology), without risking the sort of image step-change that Mercedes-Benz took in 2014 with the rival C-Class, a decision that has paid huge dividends in terms of sales and brand appreciation.
Birch, Stuart
Items per page:
1 – 50 of 157