SEARCH · Search NASA
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Automatic Speech Recognition (ASR) technology and its application to the Air Traffic Control system are described. The advantages of applying ASR to Air Traffic Control, as well as criteria for choosing a suitable ASR system are presented. Results from previous research and directions for future work at the Flight Transportation Laboratory are outlined.
In recent years, the application of automatic speech recognition has gained popularity across diverse industries, including aviation. Given the many applications focusing on transcribing air traffic control and management communication, this paper explores the training of OpenAI's Whisper model across multiple existing public and private air traffic control voice datasets in an effort to improve robustness. Combining roughly 60+ hours of various air traffic datasets, our goal is to train a unified Whisper model and expect an average word error rate reduction across testing datasets. Furthermore, this work aims to understand the data quantity requirements for achieving state-of-the art results by comprehensively training Whisper on varying dataset sizes. This work has the potential to improve automatic speech recognition performance across the domain, improve understanding of the quantity of data required by an aviation speech recognition system, and lastly provide metrics to compare and improve upon in future research.
This paper addresses the challenge of using Automatic Speech Recognition (ASR) technology to transcribe regular teleconferences that happen between FAA Air Traffic Control System Command Center (ATCSCC) planners, stakeholders and air users. These planning teleconferences (aka telecons or planning webinars) are an integral part of managing air traffic in the U.S. National Airspace System (NAS). In particular, the meetings facilitate the creation and modification of various traffic management initiatives (TMIs), that are used to regulate the flow of air traffic. This is typically a human intensive process, requiring specialists to listen to the entire meeting audio (10-20 minutes duration) and inferring the state of the NAS (e.g., weather phenomenon) that was discussed. It would be advantageous to have digital transcripts of the audio and have useful information (e.g., related to TMIs) automatically extracted from the transcripts. In this regard, we are exploring the adoption of state-of-the-art speech to text and Natural Language Processing (NLP) tools that will achieve our objective of digitizing the webinar audio. Unfortunately, the highly technical phraseology present in the audio and limited data availability for model building make ASR difficult. To overcome this challenge, we have taken the critical first step in creating a human transcription dataset from ~20 hours of speech in the ATCSCC audio with the help of subject matter experts. A novelty of our work is the creation of a ground truth transcription dataset for ATCSCC teleconference webinars, which is particularly important for Aviation domain-specific NLP tasks. Using Microsoft Speech Studio, a cloud-based ASR platform, we have fine-tuned the English pre-trained ASR models (available in speech studio) and achieved an average word error rate (WER) of 6.81%. The baseline ASR also provides a digital version of each planning webinar, making it accessible and text-searchable for future references. Additionally, the transcriptions can serve as a bridge between raw audio data and a range of text-based NLP tasks, such as named entity recognition (NER) and intent classification, potentially enhancing the digital footprint of the webinars and other connected data sources. Our work has several potential applications. Firstly, the transcriptions can be analyzed to understand the complex decision process of creating, implementing and modifying TMIs and may also contribute to TMI prediction services. Secondly, our dataset and model can be used to develop more accurate ASR systems for aviation-specific language, which can bring about digital communication in the aviation industry (and aid current “voice only” communications, which are inherently error-prone). Lastly, the transcriptions themselves can be used as a valuable resource for training other NLP models.
The decisions made during the Federal Aviation Administration Air Traffic Control System Command Center's planning teleconferences hold significant sway over the National Airspace System. Held every two hours, these teleconferences convene air traffic managers and stakeholders from across the nation to discuss airspace conditions, weather, and constraints, leading to the formulation and adjustment of traffic management initiatives. Given the critical nature of these decisions, the need for accurate and efficient record-keeping is paramount. In recent years, the application of automatic speech recognition has gained popularity across diverse industries, including aviation. While traditional applications focus on transcribing air traffic control communication, this paper explores a unique application of automatic speech recognition by converting the audio from planning teleconferences into text transcriptions. This innovative approach addresses key challenges in the field, presenting potential benefits for quality assurance, real-time participation, and downstream natural language processing tasks. A notable breakthrough in the machine learning community, namely the transformer neural network architecture, forms the backbone of the proposed solution in this paper. The transformer architecture's role in this research represents a paradigm shift in the efficiency of automatic speech recognition models. By reducing the amount of in-domain training data required, this architecture allows for the fine-tuning of such models like Whisper, originally pretrained on vast English speech datasets. The adaptability of the transformer architecture proves invaluable in capturing the nuances of aviation terminology and specific language used in planning teleconferences. Leveraging the Whisper model as a baseline, our research details the fine-tuning and validation using a dataset comprising 20 hours of meticulously transcribed planning teleconferences. Notably, the baseline pretrained Whisper model exhibited a word error rate of 18.77%. Through the fine-tuning process, the model achieved a substantial improvement, demonstrating an impressive performance with a reduced word error rate of 6.82%. This substantial decrease in WER not only highlights the effectiveness of the transformer architecture but also emphasizes the practical advancements achieved through the application of automatic speech recognition in this specific domain. The utilization of automatic speech recognition in planning teleconferences in this work introduces several novelties. Firstly, the creation of text transcriptions offers a valuable tool for quality assurance and facilitates the efficient review of teleconferences. This is an important aspect of the proposed solution, given the time-sensitive and high-stakes nature of decisions made during these meetings. Furthermore, text-searchable transcriptions provide a streamlined approach for locating and validating critical information, potentially saving hours of manual effort in searching through audio recordings. Moreover, our research identifies a key use case for external facilities and stakeholders. In situations where attendance at the planning teleconference is not feasible, having access to text transcriptions in real-time or shortly after the teleconference ends, proves to be a time-saving and informative resource. This feature enhances collaboration and ensures that stakeholders can stay abreast of important discussions and decisions even in their absence. Despite the efficiency gains facilitated by the transformer architecture in automatic speech recognition technology, it is essential to acknowledge the human factors in data creation. Subject matter experts play a crucial role in accurately transcribing planning teleconferences due to the specificity and complexity of the information discussed. The research dataset, consisting of 20 hours of transcribed planning teleconferences, forms the foundation for fine-tuning and validating the Whisper model. The achieved word error rate of 6.82% demonstrates promising advancements, particularly in recognizing essential aviation terminology within the teleconferences. In conclusion, this paper presents a comprehensive exploration of the application of automatic speech recognition in Air Traffic Control System Command Center planning teleconferences, leveraging the transformer architecture for enhanced efficiency. The novel contributions lie in the improved accessibility of decision-making records, real-time participation opportunities for external stakeholders, and the potential for downstream natural language processing advancements. As the aviation industry continues to evolve, the integration of automatic speech recognition technologies holds the promise of revolutionizing decision-making processes and contributing to the overall safety and efficiency of air traffic management.
The introduction of Automatic Speech Recognition (ASR) technology into the Air Traffic Control (ATC) system has the potential to improve overall safety and efficiency. However, because ASR technology is inherently a part of the man-machine interface between the user and the system, the human factors issues involved must be addressed. Here, some of the human factors problems are identified and related methods of investigation are presented. Research at M.I.T.'s Flight Transportation Laboratory is being conducted from a human factors perspective, focusing on intelligent parser design, presentation of feedback, error correction strategy design, and optimal choice of input modalities.
An updated summary of the state-of-the-art of automatic speech recognition and its relevance to military applications is provided. A number of potential systems for military applications are under development. These include: (1) digital narrowband communication systems; (2) automatic speech verification; (3) on-line cartographic processing unit; (4) word recognition for militarized tactical data system; and (5) voice recognition and synthesis for aircraft cockpit.
An assessment of the applications of automatic speech recognition to defense communication systems is presented. Future research efforts include investigations into the following areas: (1) dynamic programming; (2) recognition of speech degraded by noise; (3) speaker independent recognition; (4) large vocabulary recognition; (5) word spotting and continuous speech recognition; and (6) isolated word recognition.
Transcribing voice communications in NASA’s launch control center is important for information utilization. However, automatic speech recognition in this environment is particularly challenging due to the lack of training data, unfamiliar words in acronyms, multiple different speakers and accents, and conversational characteristics of speaking. We used bidirectional deep recurrent neural networks to train and test speech recognition performance. We showed that data augmentation and custom language models can improve speech recognition accuracy. Transcribing communications from the launch control center will help the machine analyze information and accelerate knowledge generation.
Abstract Significant progress has been made in probing the state of an earthquake fault by applying machine learning to continuous seismic waveforms. The breakthroughs were originally obtained from laboratory shear experiments and numerical simulations of fault shear, then successfully extended to slow-slipping faults. Here we apply the Wav2Vec-2.0 self-supervised framework for automatic speech recognition to continuous seismic signals emanating from a sequence of moderate magnitude earthquakes during the 2018 caldera collapse at the Kīlauea volcano on the island of Hawai’i. We pre-train the Wav2Vec-2.0 model using caldera seismic waveforms and augment the model architecture to predict contemporaneous surface displacement during the caldera collapse sequence, a proxy for fault displacement. We find the model displacement predictions to be excellent. The model is adapted for near-future prediction information and found hints of prediction capability, but the results are not robust. The results demonstrate that earthquake faults emit seismic signatures in a similar manner to laboratory and numerical simulation faults, and artificial intelligence models developed for encoding audio of speech may have important applications in studying active fault zones.
In the present air traffic system, information presented to the transport aircraft cockpit crew may originate from a variety of sources and may be presented to the crew in visual or aural form, either through cockpit instrument displays or, most often, through voice communication. Voice radio communications are the most error prone method for air-ground data link. Voice messages can be misstated or misunderstood and radio frequency congestion can delay or obscure important messages. To prevent proliferation, a multiplexed data link display can be designed to present information from multiple data link sources on a shared cockpit display unit (CDU) or multi-function display (MFD) or some future combination of flight management and data link information. An aural data link which incorporates an automatic speech recognition (ASR) system for crew response offers several advantages over visual displays. The possibility of applying ASR to the air-ground data link was investigated. The first step was to review current efforts in ASR applications in the cockpit and in air traffic control and evaluated their possible data line application. Next, a series of preliminary research questions is to be developed for possible future collaboration.
A voice-command human-machine interface system has been developed for spacesuit extravehicular activity (EVA) missions. A multichannel acoustic signal processing method has been created for distant speech acquisition in noisy and reverberant environments. This technology reduces noise by exploiting differences in the statistical nature of signal (i.e., speech) and noise that exists in the spatial and temporal domains. As a result, the automatic speech recognition (ASR) accuracy can be improved to the level at which crewmembers would find the speech interface useful. The developed speech human/machine interface will enable both crewmember usability and operational efficiency. It can enjoy a fast rate of data/text entry, small overall size, and can be lightweight. In addition, this design will free the hands and eyes of a suited crewmember. The system components and steps include beam forming/multi-channel noise reduction, single-channel noise reduction, speech feature extraction, feature transformation and normalization, feature compression, model adaption, ASR HMM (Hidden Markov Model) training, and ASR decoding. A state-of-the-art phoneme recognizer can obtain an accuracy rate of 65 percent when the training and testing data are free of noise. When it is used in spacesuits, the rate drops to about 33 percent. With the developed microphone array speech-processing technologies, the performance is improved and the phoneme recognition accuracy rate rises to 44 percent. The recognizer can be further improved by combining the microphone array and HMM model adaptation techniques and using speech samples collected from inside spacesuits. In addition, arithmetic complexity models for the major HMMbased ASR components were developed. They can help real-time ASR system designers select proper tasks when in the face of constraints in computational resources.
A trainable acoustic pattern recognizer manufactured by Scope Electronics is presented. The voice command system VCS encodes speech by sampling 16 bandpass filters with center frequencies in the range from 200 to 5000 Hz. Variations in speaking rate are compensated for by a compression algorithm that subdivides each utterance into eight subintervals in such a way that the amount of spectral change within each subinterval is the same. The recorded filter values within each subinterval are then reduced to a 15-bit representation, giving a 120-bit encoding for each utterance. The VCS incorporates a simple recognition algorithm that utilizes five training samples of each word in a vocabulary of up to 24 words. The recognition rate of approximately 85 percent correct for untrained speakers and 94 percent correct for trained speakers was not considered adequate for flight systems use. Therefore, the built-in recognition algorithm was disabled, and the VCS was modified to transmit 120-bit encodings to an external computer for recognition.
NASA has a widely recognized but unmet need for novel human-machine interface technologies that can facilitate communication during astronaut extravehicular activities (EVAs), when loud noises and strong reverberations inside spacesuits make communication challenging. WeVoice, Inc., has developed a multichannel signal-processing method for speech acquisition in noisy and reverberant environments that enables automatic speech recognition (ASR) technology inside spacesuits. The technology reduces noise by exploiting differences between the statistical nature of signals (i.e., speech) and noise that exists in the spatial and temporal domains. As a result, ASR accuracy can be improved to the level at which crewmembers will find the speech interface useful. System components and features include beam forming/multichannel noise reduction, single-channel noise reduction, speech feature extraction, feature transformation and normalization, feature compression, and ASR decoding. Arithmetic complexity models were developed and will help designers of real-time ASR systems select proper tasks when confronted with constraints in computational resources. In Phase I of the project, WeVoice validated the technology. The company further refined the technology in Phase II and developed a prototype for testing and use by suited astronauts.
The program at Naval Air Development Center was initiated to determine the desirability of interactive voice systems for use in airborne weapon systems crew stations. A voice recognition and synthesis system (VRAS) was developed and incorporated into a human centrifuge. The speech recognition aspect of VRAS was developed using a voice command system (VCS) developed by Scope Electronics. The speech synthesis capability was supplied by a Votrax, VS-5, speech synthesis unit built by Vocal Interface. The effects of simulated flight on automatic speech recognition were determined by repeated trials in the VRAS-equipped centrifuge. The relationship of vibration, G, O2 mask, mission duration, and cockpit temperature and voice quality was determined. The results showed that: (1) voice quality degrades after 0.5 hours with an O2 mask; (2) voice quality degrades under high vibration; and (3) voice quality degrades under high levels of G. The voice quality studies are summarized. These results were obtained with a baseline of 80 percent recognition accuracy with VCS.
The pilot's perception and performance in flight simulators is examined. The areas investigated include: vestibular stimulation, flight management and man cockpit information interfacing, and visual perception in flight simulation. The effects of higher levels of rotary acceleration on response time to constant acceleration, tracking performance, and thresholds for angular acceleration are examined. Areas of flight management examined are cockpit display of traffic information, work load, synthetic speech call outs during the landing phase of flight, perceptual factors in the use of a microwave landing system, automatic speech recognition, automation of aircraft operation, and total simulation of flight training.
Automatic Speech Recognition technology has matured to the point where it can provide a viable means of increasing productivity by naturalizing the man-machine interface. With ever increasing workloads being placed on astronauts, speech recognition may provide an alternative means of system controlling that would reduce the task burden. Voice commanding, allowing hands-free operation, can be especially effective during operations requiring simultaneous system control. A flight experiment is under development to demonstrate the operational effectiveness of voice control by commanding the Space Shuttle's Closed Circuit Television (CCIV) system. This experiment will help direct future applications of voice entry to space operations.
There are four areas of research being pursued in 1988 under sponsorship of the FAA/NASA Joint University Research Program, and one area which has been completed. The four active areas were: (1) Automatic Speech Recognition (ASR) in Air Traffic Control. The purpose of this research effort is to demonstrate the feasibility of using ASR technology within the ATC environment and to address the problems involved, especially the relevant human factors issues. (2) A Rule-Based Planning and Scheduling System. Planning denotes the formulation of a detailed scheme, program, or method worked out beforehand for the accomplishment of a goal. It involves the analysis of the desired goal and its division into sub-goals which are subsequently treated in the same way until a set of primitive objectives is obtained. (3) Modeling of Ice Accretion on Aircraft in Glaze Icing Conditions. The work in aircraft icing over the past year has focused on the fundamental aspects of glaze ice accretion, with the goal of improving analytical ice accretion models. Over the past year, studies have been conducted on the generation of surface roughness on accreting ice surfaces with the goal of providing a deterministic surface roughness in the ice accretion models. (4) Cockpit Display of Hazardous Weather Information. Information transfer and display issues associated with the dissemination of hazardous weather warnings are studied in the context of windshear alerts.
The following research was discussed: (1) speech signal processing; (2) automatic speech recognition; (3) continuous speech understanding; (4) speaker recognition; (5) speech compression; (6) subjective and objective evaluation of speech communication system; (7) measurement of the intelligibility and quality of speech when degraded by noise or other masking stimuli; (8) speech synthesis; (9) instructional aids for second-language learning and for training of the deaf; and (10) investigation of speech correlates of psychological stress. Experimental psychology, control systems, and human factors engineering, which are often relevant to the proper design and operation of speech systems are described.