Run-Time Safety Monitoring Framework for AI-Based Systems: Automated Driving Cases

Osman, Mohd Hafeez; Kugele, Stefan; Shafaei, Sina

doi:10.1109/apsec48747.2019.00066

Cited by 5 publications

(2 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Publications within this field mostly concentrate on the technological feasibility of meeting the requirements for intelligent, safe and stable driving. Researchers have developed and applied recognition systems for an autonomous vehicle system (Choi et al , 2019), pedestrian detection (Karg and Scharfenberger, 2020) and run-time safety monitoring (Osman et al , 2019). Third, AI has taken the form of onboard predictive maintenance components, which are applied to handle the equipment management of trucks by detecting the maintenance cycle and instances of maintenance based on big data (Sun et al , 2019).…”

Section: Theoretical Background and Underpinning Theoriesmentioning

confidence: 99%

Intelligent and efficient? An empirical analysis of human–AI collaboration for truck drivers in retail logistics

Loske

Klumpp

2021

IJLM

View full text Add to dashboard Cite

PurposeTechnological advances regarding artificial intelligence (AI) are affecting the transport sector. Although fully autonomous delivery, or self-driving trucks, are not operating currently, various AI applications have become fixed components of cargo vehicles. Since many research approaches primarily concentrate on the technical aspects of assistance systems (ASs), the economic question of how to improve efficiency is seldom addressed. Therefore, the purpose of this paper is to apply an efficiency analysis to measure the performance of truck drivers supplying retail stores.Design/methodology/approachFor this comparative study, 90 professional truck drivers in three groups are compared with (1) trucks without AS, (2) trucks with AS that cannot be turned off and (3) trucks with AS that can be turned off. First, we build a model investigating the impact of performance expectation, effort expectation, social influence and facilitating conditions on the behavioural intention to use AS. Second, we explore the impact of truck drivers' behavioural intention on actual technology use, misuse and disuse; operationalize these constructs; and merge them with our behavioural constructs to create one econometric model.FindingsThe human–AI system was found to be the most efficient. Additionally, behavioural intention to use ASs did not lead to actual usage in the AI-alone observation group, but did in the human–AI group. Several in-depth analyses showed that the AI-alone group used AS at a higher level than the human–AI group, but manipulations through, for example, kickdowns or manual break operations led to conscious overriding of the cruise control system and, consequently, to higher diesel consumption, higher variable costs and lower efficiency of transport logistical operations.Research limitations/implicationsEfficiency analysis with data envelopment analysis is, by design, limited by the applied input and output factors.Originality/valueThis study represents one of the first quantitative efficiency analyses of the impact of digitalization on transport performance (i.e. truck driver efficiency). Furthermore, we build an econometric model combining behavioural aspects with actual technology usage in a real application scenario.

show abstract

Section: Theoretical Background and Underpinning Theoriesmentioning

confidence: 99%

Intelligent and efficient? An empirical analysis of human–AI collaboration for truck drivers in retail logistics

Loske

Klumpp

2021

IJLM

View full text Add to dashboard Cite

show abstract

“…These autonomous AEVs redefine the vehicle's business environment as the gigantic information generated by the vehicles themselves. Aggressors have shifted their focus towards the hack of information associated/generated with/by autonomous AEVs, which lead them to control the entire vehicle and incapacitate it also (Habeeb et al, 2019;Osman et al, 2019). Table 2 present various security vulnerabilities associated with specific EV components along with possible AI techniques to minimize its effect.…”

mentioning

confidence: 99%

Anomaly detection in autonomous electric vehicles using AI techniques: A comprehensive survey

et al. 2021

View full text Add to dashboard Cite

The next wave in smart transportation is directed towards the design of renewable energy sources that can fuel automobile sector to shift towards the autonomous electric vehicles (AEVs). AEVs are sensor-driven and driverless that uses artificial intelligence (AI)-based interactions in Internet-of-vehicles (IoV) ecosystems. AEVs can reduce carbon footprints and trade energy with peer AEVs, smart grids (SG), and roadside units (RSUs). It supports green transportation vision. However, the sensor information, energy units, and user data are exchanged through open channels, and thus, are susceptible to various security and privacy attacks. Thus, AEVs can be remotely operated and directed by malicious entities that can propagate false updates to the peer nodes in IoV environment. This can cause the failure of components, congestion, as well as the entire disruption of IoV network. Globally researchers and security analysts have addressed solutions that pertain to specific security requirements, but still, the detection and classification of malicious AEVs is a widely studied topic. Malicious AEVs exhibit an anomaly behavior that differentiates them from normal AEVs, and thereby, the detection of anomalous AEVs and classification of anomaly type is required. Motivated from the aforementioned facts, the survey presents a systematic outlook of AI techniques in anomaly detection of AEVs.A solution taxonomy is proposed based on research gaps in the existing surveys, and the evaluation metrics for AI-based anomaly detection are discussed. The open challenges and issues in AI deployments are discussed and a case study is presented on anomaly classification through a weighted ensemble technique. Thus, the proposed survey is designed to guide the manufacturing industry, AI practitioners, and researchers worldwide to formulate and design accurate and precise mechanisms to detect anomalies.

show abstract

Assuring the Safety of End-to-End Learning-Based Autonomous Driving through Runtime Monitoring

Grieser

Zhang

Warnecke

et al. 2020

2020 23rd Euromicro Conference on Digital System Design (DSD)

View full text Add to dashboard Cite

Run-Time Safety Monitoring Framework for AI-Based Systems: Automated Driving Cases

Cited by 5 publications

References 12 publications

Intelligent and efficient? An empirical analysis of human–AI collaboration for truck drivers in retail logistics

Intelligent and efficient? An empirical analysis of human–AI collaboration for truck drivers in retail logistics

Anomaly detection in autonomous electric vehicles using AI techniques: A comprehensive survey

Assuring the Safety of End-to-End Learning-Based Autonomous Driving through Runtime Monitoring

Contact Info

Product

Resources

About