Toward AI-enabled augmented reality to enhance the safety of highway work zones: Feasibility,requirements, and challenges

Highway work zones are considered among the most hazardous working environments. In 2018 alone, 124 workers lost their lives to fatal accidents. The lack of predictive safety systems that notify workers of upcoming dangers in advance is a major reason to blame in the highway maintenance and operation community. This article presents an integrative design framework for bringing recent advances in Augmented Reality (AR) and Artificial Intelligence (AI) to enhance the safety of highway workers through real-time multimodal notifications on-spot. To this end, this article conceptualizes and co-designs three major pillars: (1) AR user interface design for multimodal notification, (2) real-time AI at the edge for vehicle detection/classification from distance, and (3) real-time wireless communication in work zone setting to enable latency-aware operation between AI and AR components. Our early results demonstrate that we can achieve 24.83 FPS end-to-end execution latency on the Xavier AGX Jetson board with 48.7% mAP on BDD100K dataset, and a real-time communication covering 120 meters with an average latency of 5.1 ms at the farthest distance. Our mixed-method user research also reveals an acceptable level of excitement and engagement from the body of highway workers toward both the proposed technology and the designed user interface. Overall, this article provides a proof-of-concept toward AI-enabled AR safety systems in highway work zones.     

Parallel Knowledge Representation on the Connection Machine

Semantic network systems are a common knowledge representation paradigm used in contemporary AI research. Researchers have become increasingly discouraged, however, with the performance of these systems. As networks grow performance suffers, often to an unacceptable extent. The representations for many AI domains are so large that existing systems cannot execute general inferences on them in an acceptable amount of time-serial representation systems are computationally ineffective for large domains. This is one of the reasons that, in the last few years, researchers have been turning their attention, away from traditional symbolic approaches, toward neural networks and parallel connectionist paradigms. We have been developing PARKA, a symbolic, semantic network knowledge representation system that takes advantage of the CM′s massive parallelism. Our primary motivation is to demonstrate that massive parallelism can be used to provide the run-time improvements necessary to bring large AI systems into realistic applications requiring rapid response time. PARKA′s design allows it to evaluate rapidly enough for everyday use many inferences that are computationally ineffective in serial systems. In this paper we describe the design and implementation of PARKA. We explain our motivations for developing a representation system on the Connection Machine and show how PARKA exploits its design to optimize performance. We show that PARKA effects in O(d) time inferences for which serial systems require O(B^d) time, where d is the network depth and B is the average network branchout. Timings of the system on random networks are also presented. We show that PARKA is capable of effecting top-down inheritance inferences on 16,000-node networks in well under a second.     

实时人工智能的研究

1 引言随着实时技术的发展,实时系统对制造、控制、运输、太空、机器人和军事系统起着越来越关键性的作用。实时系统是工作在时间约束下的系统,它与一般的计算机系统有本质的区别。实时系统不但要保证计算结果的逻辑正确性,而且必需在截止期内完成任务。在硬实时系统中,如果实时任务没有在规定的截止期内     

Contextual influences on self-control of is professionals engaged in systems development

In today's turbulent business environment, there is a move away from traditional hierarchical relations and governance within organizations, and a move toward increased reliance on self-control, where an individual sets his own goals, monitors his own work, and rewards or sanctions himself accordingly. Since the use of self control is often recommended in an environment of task complexity and ambiguity, systems development would seem to provide an excellent context in which to study it. The goal of this research is to examine contextual factors (work unit structure and knowledge technology) that influence IS project leaders' perceptions of self-control. To meet this goal, two studies were carried out: a survey of IS professionals and a series of three case studies of systems development efforts. Overall, the results suggest that IS project leaders' perceptions of self-control are highest when they have considerable job experience, when they are able to further refine existing development procedures, and when they are involved in smaller, less-complex systems development projects. Implications for the practice of systems development are discussed.     

Temporal Reasoning for a Collaborative Planning Agent in a Dynamic Environment

We present a temporal reasoning mechanism for an individual agent situated in a dynamic environment such as the web and collaborating with other agents while interleaving planning and acting. Building a collaborative agent that can flexibly achieve its goals in changing environments requires a blending of real-time computing and AI technologies. Therefore, our mechanism consists of an Artificial Intelligence (AI) planning subsystem and a Real-Time (RT) scheduling subsystem. The AI planning subsystem is based on a model for collaborative planning. The AI planning subsystem generates a partial order plan dynamically. During the planning it sends the RT scheduling subsystem basic actions and time constraints. The RT scheduling subsystem receives the dynamic basic actions set with associated temporal constraints and inserts these actions into the agent's schedule of activities in such a way that the resulting schedule is feasible and satisfies the temporal constraints. Our mechanism allows the agent to construct its individual schedule independently. The mechanism handles various types of temporal constraints arising from individual activities and its collaborators. In contrast to other works on scheduling in planning systems which are either not appropriate for uncertain and dynamic environments or cannot be expanded for use in multi-agent systems, our mechanism enables the individual agent to determine the time of its activities in uncertain situations and to easily integrate its activities with the activities of other agents. We have proved that under certain conditions temporal reasoning mechanism of the AI planning subsystem is sound and complete. We show the results of several experiments on the system. The results demonstrate that interleave planning and acting in our environment is crucial.     

Artificial intelligence and natural magic

Robotics with AI is part of a long tradition that has run from ancient times that treated the precursors of robots, the automata, as part of Natural Magic or conjury. Deception is an integral part of AI and robotics; in some ways they form a science of illusion. There are many robot tasks, such as caring for the elderly, minding children, doing domestic chores and being companionable, that involve working closely with humans and so require some illusion of animacy and thought. We discuss how the natural magic of robotics is assisted by the cultural myth of AI together with innate human predispositions such as zoomorphism, the willing suspension of disbelief and a tendency to interpret AI devices as part of the social world. This approach provides a justifiable way of meeting the goals of AI and robotics provided that researchers do not allow themselves to be deceived by their own illusions.     

Analysis of real-time rule-based systems with behavioral constraintassertions specified in Estella

Rule-based expert systems are increasingly used to monitor and control the operations of complex real-time systems which require intensive knowledge-decision processing and human expertise. These embedded AI systems must respond to events in the rapidly changing external environment so that the results of the expert system's computation in each monitor-respond cycle are valid in safely operating the real-time system. Determining how fast an expert system can respond under all possible situations is a difficult problem. We have developed an efficient analysis methodology for a large class of rule-based EQL programs to determine whether a program in this class has bounded response time. In particular, we have identified several sets of primitive behavioral constraint assertions: an EQL program which satisfies all constraints in one of these sets of assertions is guaranteed to have bounded response time. Here, we enhance the applicability of our analysis technique by introducing a facility with which the rule-based programmer can specify application-specific knowledge that is too difficult to be mechanically detected in the new language Estella in order to determine the performance of an even wider range of programs. We also describe efficient algorithms for implementing the analysis tools     

Trends in distributed artificial intelligence

Distributed artificial intelligence (DAI) is a subfield of artificial intelligence that deals with interactions of intelligent agents. Precisely, DAI attempts to construct intelligent agents that make decisions that allow them to achieve their goals in a world populated by other intelligent agents with their own goals. This paper discusses major concepts used in DAI today. To do this, a taxonomy of DAI is presented, based on the social abilities of an individual agent, the organization of agents, and the dynamics of this organization through time. Social abilities are characterized by the reasoning about other agents and the assessment of a distributed situation. Organization depends on the degree of cooperation and on the paradigm of communication. Finally, the dynamics of organization is characterized by the global coherence of the group and the coordination between agents. A reasonably representative review of recent work done in DAI field is also supplied in order to provide a better appreciation of this vibrant AI field. The paper concludes with important issues in which further research in DAI is needed.     

Cancer diagnosis using artificial intelligence: a review

Artificial intelligence (AI) is the usage of scientific techniques to simulate human intellectual skills and to tackle complex medical issues involving complicated genetic defects such as cancer. The rapid expansion of AI in the past era has paved the way to optimum judgment-making by superior intellect, where the human brain is constrained to manage large information in a limited period. Cancer is a complicated ailment along with several genomic variants. AI-centred systems carry enormous potential in detecting these genomic alterations and abnormal protein communications at a very initial phase. The contemporary biomedical study is also dedicated to bringing AI expertise to hospitals securely and ethically. AI-centred support to diagnosticians and doctors can be the big surge ahead for the forecast of illness threat, identification, diagnosis, and therapies. The applications related to AI and Machine Learning (ML) in the identification of cancer and its therapy possess the potential to provide therapeutic support for quicker planning of a novel therapy for each person. Through the utilization of AI- based methods, scientists can work together in real-time and distribute their expertise digitally to possibly cure billions. In this review, the focus was on the study of linking biology with AI and describe how AI-centred support could assist oncologists in accurate therapy. It is essential to identify new biomarkers that inject drug defiance and discover medicinal goals to improve medication methods. The advent of the “next-generation sequencing” (NGS) programs resolves these challenges and has transformed the prospect of “Precision Oncology” (PO). NGS delivers numerous medical functions which are vital for hazard prediction, initial diagnosis of infection, “Sequence” identification and “Medical Imaging” (MI), precise diagnosis, “biomarker” detection, and recognition of medicinal goals for innovation in medicine. NGS creates a huge repository that requires specific “bioinformatics” sources to examine the information that is pertinent and medically important. The malignancy diagnostics and analytical forecast are improved with NGS and MI that provide superior quality images via AI technology. Irrespective of the advancements in technology, AI faces a few problems and constraints, and the clinical application of NGS continues to be authenticated. Through the steady progress of invention and expertise, the prospects of AI and PO look promising. The purpose of this review was to assess, evaluate, classify, and tackle the present developments in cancer diagnosis utilizing AI methods for breast, lung, liver, skin cancer, and leukaemia. The research emphasizes in what way cancer identification, the treatment procedure is aided by utilizing AI with supervised, unsupervised, and deep learning (DL) methods. Numerous AI methods were assessed on benchmark datasets with respect to “accuracy”, “sensitivity”, “specificity”, and “false-positive” (FP) metrics. Lastly, challenges along with future work were discussed.

     

Machine intelligence-the animat path to intelligent adaptivebehaviour

The animat path to artificial intelligence (AI), a bottom-up approach to creating intelligent systems, is described. Using this approach, artificial creatures or agents animats are constructed in an environment. They begin simply and are gradually made more and more complex, exhibiting more and more complex behaviors at each step. The goals and accomplishments of five research projects incorporating the animat path to AI methodology are reviewed     

The inner and the outer model in explanatory design theory: the case of designing electronic feedback systems

Both Information Technology (IT) artifacts and design theories are important elements for knowledge capture in design science research in information systems. Building on a rich tradition of constructing and evaluating artifacts, recent design science research has made significant advances toward better understanding the explanatory aspect of design theory. Researchers have stressed the importance of mid-range theories that relate IT artifact features (causes) with measures and goals (effects). Against this background, design theorizing reveals certain commonalities with theorizing in the behavioral science field. In this paper, we explore differences and similarities between theorizing in these areas. We develop a framework that allows for a better understanding of the relationships between manifest design decisions, kernel theory constructs and their evaluation metrics. We identify common issues that arise from conceptual distances between these ideas and show their potential impact on both the design and evaluation of artifacts. The field of electronic feedback systems is used as an illustrative example.     

Intelligent computer-aided-design systems: a synergical approach of artificial intelligénce and engineering

A synergism has begun to surface from the artificial intelligence (AI) and engineering communities: an effort to apply AI techniques to engineering problem-solving activities, and to study problems arisen from various engineering fields as a way to develop AI theories and methodologies. This paper first discusses the needs of such a synergical approach and identifies in a broad perspective some AI techniques currently being applied to engineering. It then describes a system, called KREATOR, which applies qualitative reasoning, a subfield of AI, to computer-aided design (CAD). The key observation is that an engineer designer's qualitative knowledge can offer a good basis for the reasoning of device behaviors. Such knowledge, however, is not captured by conventional CAD systems for lack of good representations. KREATOR is a knowledge capturing scheme that allows the designers to record their qualitative knowledge of how mechanical devices behave, KREATOR then automatically generates qualitative simulations.     

Layered approach to learning client behaviors in the robocup soccer server

In the past few years, multiagent systems (MAS) have emerged as an active subfield of artificial intelligence (AI). Because of the inherent complexity of MAS, there is much interest in using machine learning (ML) techniques to help build multiagent systems. Robotic soccer is a particularly good domain for studying MAS and multiagent learning. Our approach to using ML as a tool for building Soccer Server clients involves layering increasingly complex learned behaviors. In this article, we describe two levels of learned behaviors. First, the clients learn a low-level individual skill that allows them to control the ball effectively. Then, using this learned skill, they learn a higher level skill that involves multiple players. For both skills, we describe the learning method in detail and report on our extensive empirical testing. We also verify empirically that the learned skills are applicable to game situations.     

On the reliability of AI planning software in real-timeapplications

We define the reliability of a real-time system incorporating AI planning programs as the probability that, for each problem-solving request issued from the environment, the embedded system can successfully plan and execute a response within a specified real-time deadline. A methodology is developed for evaluating the reliability of such systems taking into consideration the fact that, other than program bugs, the intrinsic characteristics of AI planning programs may also cause the embedded system to fail even after all software bugs are removed from the program. The utility of the methodology is demonstrated by applying it to the reliability evaluation of two AI planning algorithms embedded in a real-time multicriteria route finding system     

Utility Models for Goal‐Directed, Decision‐Theoretic Planners

AI planning agents are goal-directed : success is measured in terms of whether an input goal is satisfied. The goal gives structure to the planning problem, and planning representations and algorithms have been designed to exploit that structure. Strict goal satisfaction may be an unacceptably restrictive measure of good behavior, however.
A general decision-theoretic agent, on the other hand, has no explicit goals: success is measured in terms of an arbitrary preference model or utility function defined over plan outcomes. Although it is a very general and powerful model of problem solving, decision-theoretic choice lacks structure, which can make it difficult to develop effective plan‐generation algorithms.
This paper establishes a middle ground between the two models. We extend the traditional AI goal model in several directions: allowing goals with temporal extent, expressing preferences over partial satisfaction of goals, and balancing goal satisfaction against the cost of the resources consumed in service of the goals. In doing so we provide a utility model for a goal-directed agent.
An important quality of the proposed model is its tractability. We claim that our model, like classical goal models, makes problem structure explicit. This structure can then be exploited by a problem-solving algorithm. We support this claim by reporting on two implemented planning systems that adopt and exploit our model.     

Multiagent Systems: A Survey from a Machine Learning Perspective

Distributed Artificial Intelligence (DAI) has existed as a subfield of AI for less than two decades. DAI is concerned with systems that consist of multiple independent entities that interact in a domain. Traditionally, DAI has been divided into two sub-disciplines: Distributed Problem Solving (DPS) focuses on the information management aspects of systems with several components working together towards a common goal; Multiagent Systems (MAS) deals with behavior management in collections of several independent entities, or agents. This survey of MAS is intended to serve as an introduction to the field and as an organizational framework. A series of general multiagent scenarios are presented. For each scenario, the issues that arise are described along with a sampling of the techniques that exist to deal with them. The presented techniques are not exhaustive, but they highlight how multiagent systems can be and have been used to build complex systems. When options exist, the techniques presented are biased towards machine learning approaches. Additional opportunities for applying machine learning to MAS are highlighted and robotic soccer is presented as an appropriate test bed for MAS. This survey does not focus exclusively on robotic systems. However, we believe that much of the prior research in non-robotic MAS is relevant to robotic MAS, and we explicitly discuss several robotic MAS, including all of those presented in this issue.     

Trends in artificial intelligence applications for real-time control (a speculative study)

Some of the theory-bound evolutional trends in real-time AI applications are pointed out based on analysis of essential properties of real-time systems as well as AI based systems. The evolutional mainstream is increasing interdisciplinary integration. Three subtrends are illustrated on examples: mechanical combination of methods, AI methods used for approximate solution of “classical” problems, and abstract methods applied in new domains. In addition similarity between integrated circuits and real-time systems design and increased use of formal verification at the early stages of systems development are pointed out.     

A systematic literature review on the detection of smells and their evolution in object-oriented and service-oriented systems

This systematic literature review paper investigates the key techniques employed to identify smells in different paradigms of software engineering from object-oriented (OO) to service-oriented (SO). In this review, we want to identify commonalities and differences in the identification of smells in OO and SO systems. Our research method relies on an automatic search from the relevant digital libraries to find the studies published since January 2000 on smells until December 2017. We have conducted a pilot and author-based search that allows us to select the 78 most relevant studies after applying inclusion and exclusion criteria. We evaluated the studies based on the smell detection techniques and the evolution of different methodologies in OO and SO. Among the 78 relevant studies selected, we have identified six different studies in which linguistic source code analysis received less attention from the researchers as compared to the static source code analysis. Smells like the yo-yo problem, unnamed coupling, intensive coupling, and interface bloat received considerably less attention in the literature. We also identified a catalog of 30 smells infrequently reported for SO systems and that require further attention. Moreover, a suite of 20 smells reported for SO systems can also be detected using static source code metrics in OO. Finally, our review highlighted three major research trends that are further subdivided into 20 research patterns initiating the detection of smells toward their correction.     

ARTIFICIAL INTELLIGENCE GOES MOBILE

New mobile computing technologies require new paradigms for infrastructure and interaction with mobile and networked devices. For building smart mobile companions for new intelligent services, a number of challenges have to be addressed. We argue that artificial intelligence is a key to a new generation of mobile systems. In this introduction to AI in mobile systems, we present some of the challenges and solutions in this exciting field of research.