Identifying and managing risks of AI-driven operations: A case study of automatic speech recognition for improving air traffic safety

Lin, Yi; Ruan, Min; Cai, Kunjie; Li, Dan; Zeng, Ziqiang; Li, Fan; Yang, Bo

doi:10.1016/j.cja.2022.08.020

Cited by 22 publications

(6 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In conclusion, Generative Artificial Intelligence (AI) has the potential to transform the air civil aviation sector by providing new and innovative solutions for various applications such as flight optimization (Wu and et al, 2022), predictive maintenance (Zeng and et al, 2022), traffic management (Schweiger and et al, 2021), customer service (Ntintakis & Stavroulakis, 2020) and environmental monitoring (Yi and et al, 2022). The ability of generative AI to create new and original content, as opposed to simply recognizing or classifying existing content, opens up a wide range of possibilities for improving the efficiency, safety, and sustainability of air travel.…”

Section: Discussionmentioning

confidence: 99%

The Design of Mechanisms via Artificial Intelligence

Denizhan¹

2023

Interdisciplinary Studies on Contemporary Research Practices in Engineering in the 21st Century II

View full text Add to dashboard Cite

The engineering design is a very comprehensive term and required a number of different disciplines work together. There are many examples of engineering design studies and design of mechanisms is one of them. The design of mechanisms is one of the focused areas by mechanical engineering field and it includes kinematic synthesis, kinematic analysis, kinetics, dynamics, force and optimization of the mechanisms. The mechanisms can be complex; hence, additional help is needed in solution process. At that point, artificial intelligence (AI) of computer science is a really helpful for the mechanical engineers. Artificial intelligence field was found in 1956 and it is applicable to significant parts of the mechanism design. However, mechanical engineers cannot really apply artificial intelligence principles to mechanism design and as a result, design of mechanisms via AI became an interdisciplinary engineering studies between mechanical engineering and computer engineering. In this chapter of the book, mechanism design with artificial intelligence is focused and some sample applications are shown.

show abstract

Section: Discussionmentioning

confidence: 99%

The Design of Mechanisms via Artificial Intelligence

Denizhan¹

2023

Interdisciplinary Studies on Contemporary Research Practices in Engineering in the 21st Century II

View full text Add to dashboard Cite

show abstract

“…The later shows that novel data-driven machine learning approaches enable costly adaptations to different airport environments [2]. Lin [3,4] reviews ten tasks on spoken instruction understanding of air traffic control (ATC) data. Semisupervised learning has also been explored on the framework of ATC [5].…”

Section: Speaker Labelmentioning

confidence: 99%

“…However, in ATC communications, given its limitations such as high speaker rate, close-talk, and noise levels, relying solely on the acoustic level has shown to be insufficient. Additionally, standard SD systems add one layer of complexity to the whole ATC pipeline, 4 weakening the flexibility to transfer the already tuned pipelines to other environments (e.g., noise level variation or new speakers' accents). That is why applying SD solely on the text level stands as an interesting solution to target these disadvantages.…”

Section: Motivationmentioning

confidence: 99%

Bertraffic: Bert-Based Joint Speaker Role and Speaker Change Detection for Air Traffic Control Communications

Zuluaga-Gómez

Sarfjoo

Prasad

et al. 2023

2022 IEEE Spoken Language Technology Workshop (SLT)

View full text Add to dashboard Cite

Automatic speech recognition (ASR) allows transcribing the communications between air traffic controllers (ATCOs) and aircraft pilots. The transcriptions are used later to extract ATC named entities, e.g., aircraft callsigns. One common challenge is speech activity detection (SAD) and speaker diarization (SD). In the failure condition, two or more segments remain in the same recording, jeopardizing the overall performance. We propose a system that combines SAD and a BERT model to perform speaker change detection and speaker role detection (SRD) by chunking ASR transcripts, i.e., SD with a defined number of speakers together with SRD. The proposed model is evaluated on real-life public ATC databases. Our BERT SD model baseline reaches up to 10% and 20% token-based Jaccard error rate (JER) in public and private ATC databases. We also achieved relative improvements of 32% and 7.7% in JERs and SD error rate (DER), respectively, compared to VBx, a well-known SD system. 1

show abstract

“…Karlsson et al previously hypothesized in 1990 on this basis that the introduction of ASR technology into ATC could result in a reduced occurrence of human-generated errors enabling in turn increased safety of the overall system [17]. More recently in 2023, the use of ASR as a safety enhancing application in ATC operations was investigated and noted that the solutions investigated can improve the safety of ATC operations and can contribute to the reduction in ATCO workload [18]. Zhou et al argued that ASR represents a gateway between the ATM system and the ATCO in converting speech signal to text inputs and that after spoken instruction understanding (SIU) is applied to the converted text the output information can be used to support safety-critical applications (SCA), enabling safety and reducing possible human errors [19].…”

Section: Introductionmentioning

confidence: 99%

Ensuring Safety for Artificial-Intelligence-Based Automatic Speech Recognition in Air Traffic Control Environment

Pinska-Chauvin,

Helmke,

Dokic

et al. 2023

Aerospace

View full text Add to dashboard Cite

This paper describes the safety assessment conducted in SESAR2020 project PJ.10-W2-96 ASR on automatic speech recognition (ASR) technology implemented for air traffic control (ATC) centers. ASR already now enables the automatic recognition of aircraft callsigns and various ATC commands including command types based on controller–pilot voice communications for presentation at the controller working position. The presented safety assessment process consists of defining design requirements for ASR technology application in normal, abnormal, and degraded modes of ATC operations. A total of eight functional hazards were identified based on the analysis of four use cases. The safety assessment was supported by top-down and bottom-up modelling and analysis of the causes of hazards to derive system design requirements for the purposes of mitigating the hazards. Assessment of achieving the specified design requirements was supported by evidence generated from two real-time simulations with pre-industrial ASR prototypes in approach and en-route operational environments. The simulations, focusing especially on the safety aspects of ASR application, also validated the hypotheses that ASR reduces controllers’ workload and increases situational awareness. The missing validation element, i.e., an analysis of the safety effects of ASR in ATC, is the focus of this paper. As a result of the safety assessment activities, mitigations were derived for each hazard, demonstrating that the use of ASR does not increase safety risks and is, therefore, ready for industrialization.

show abstract

Identifying and managing risks of AI-driven operations: A case study of automatic speech recognition for improving air traffic safety

Cited by 22 publications

References 48 publications

The Design of Mechanisms via Artificial Intelligence

The Design of Mechanisms via Artificial Intelligence

Bertraffic: Bert-Based Joint Speaker Role and Speaker Change Detection for Air Traffic Control Communications

Ensuring Safety for Artificial-Intelligence-Based Automatic Speech Recognition in Air Traffic Control Environment

Contact Info

Product

Resources

About