Context awareness adds intelligence to and enriches data for applications, services and systems while enabling underlying algorithms to sense dynamic changes in incoming data streams. Context-aware machine learning is often adopted in intelligent services by endowing meaning to Internet of Things(IoT)/ubiquitous data. Intelligent transportation systems (ITS) are at the forefront of applying context awareness with marked success. In contrast to non-context-aware machine learning models, context-aware machine learning models often perform better in traffic prediction/classification and are capable of supporting complex and more intelligent ITS decision-making. This paper presents a comprehensive review of recent studies in context-aware machine learning for intelligent transportation, especially focusing on road transportation systems. State-of-the-art techniques are discussed from several perspectives, including contextual data (e.g., location, time, weather, road condition and events), applications (i.e., traffic prediction and decision making), modes (i.e., specialised and general), learning methods (e.g., supervised, unsupervised, semi-supervised and transfer learning). Two main frameworks of context-aware machine learning models are summarised. In addition, open challenges and future research directions of developing context-aware machine learning models for ITS are discussed, and a novel context-aware machine learning layered engine (CAMILLE) architecture is proposed as a potential solution to address identified gaps in the studied body of knowledge.

The Internet of Things (IoT) is growing at a rapid pace. Applications using IoT technologies have transformed many activities to be digitized enabling more productivity, economy, and quality of work. Context Management Platforms (CMPs) that unify heterogeneous streams of big IoT data and derive insights of a sensed environment (called context) using inferencing, massively enhance the smartness of IoT-based applications. Handling this massive scale of data and processing in an IoT-ecosystem for context-aware applications are both time and resource consuming, especially consid-ering the bottlenecks in network and processing resources. While traditional data caching is a time-proven technique to enable low-latency delivery of data items for a superior perceived user experience, work in context caching is extremely limited. Context information is different from typically discussed forms of data in many ways. Context-aware traditional data caching techniques have limited applicability to caching context information due to many unique challenges. These challenges can be categorized by features of context, context quality demands, techniques for caching con-text information, and context cache memory technologies. We categorically discuss each challenge in this article supported by real-world scenarios and experimental results. We contend that context caching is distinct from traditional data caching techniques and highlight the importance of con-text caching for time-critical, adaptive, context-aware applications. This article aims to demystify the unique research opportunities and challenges when developing context caching techniques for scale and efficiency objectives of CMPs and provide directions for future work.

Making internet-of-things (IoT)-based applications context-aware demands large amounts of raw data to be collected, interpreted, stored, and reused or repurposed if needed from many domains and applications. Context is transient but interpreted data can be distinguished from IoT data in many aspects. Managing context in cache is a novel area of research that has been given very little attention. Performance metric-driven adaptive context caching (ACOCA) can have a profound impact on the performance and cost efficiency of context-management platforms (CMPs) when responding to context queries in realtime. Our paper proposes an ACOCA mechanism to maximize both the cost and performance efficiency of a CMP in near realtime. Our novel mechanism encompasses the entire context-management life cycle. This, in turn, distinctively addresses the problems of efficiently selecting context for caching and managing the additional costs of context management in the cache. We demonstrate that our mechanism results in long-term efficiencies for the CMP that have not been observed in any previous study. The mechanism employs a novel, scalable, and selective context-caching agent implemented using the twin delayed deep deterministic policy gradient method. It further incorporates an adaptive context-refresh switching policy, a time-aware eviction policy, and a latent caching decision management policy. We point out in our findings that the additional complexity of adaptation introduced to the CMP through ACOCA is significantly justified, considering the cost and performance gains achieved. Our algorithm is evaluated using a real-world inspired heterogeneous context-query load and a data set based on parking-related traffic in Melbourne, Australia. This paper presents and benchmarks the proposed scheme against traditional and context-aware caching policies. We demonstrate that ACOCA outperforms the benchmarks in both cost and performance efficiency, i.e., up to 68.6%, 84.7%, and 67% more cost efficient compared to traditional data caching policies to cache context, redirector mode, and context-aware adaptive data caching under real-world-like circumstances.

The rapid growth in Internet of Things (IoT) has ushered in the way for better context-awareness enabling more smarter applications. Although for the growth in the number of IoT devices, Context Management Platforms (CMPs) that integrate different domains of IoT to produce context information lacks scalability to cater to a high volume of context queries. Research in scalability and adaptation in CMPs are of significant importance due to this reason. However, there is limited methods to benchmarks and validate research in this area due to the lack of sizable sets of context queries that could simulate real-world situations, scenarios, and scenes. Commercially collected context query logs are not publicly accessible and deploying IoT devices, and context consumers in the real-world at scale is expensive and consumes a significant effort and time. Therefore, there is a need to develop a method to reliably generate and simulate context query loads that resembles real-world scenarios to test CMPs for scale. In this paper, we propose a context query simulator for the context-aware smart car parking scenario in Melbourne Central Business District in Australia. We present the process of generating context queries using multiple real-world datasets and publicly accessible reports, followed by the context query execution process. The context query generator matches the popularity of places with the different profiles of commuters, preferences, and traffic variations to produce a dataset of context query templates containing 898,050 records. The simulator is executable over a seven-day profile which far exceeds the simulation time of any IoT system simulator. The context query generation process is also generic and context query language independent.

Context caching plays an increasingly important role in delivering near real-time responses for context-aware distributed Internet of Things (IoT) applications, services and systems. A context management platform (CMP), a middleware which acts as an aggregator and redirector of contextual information to support smart IoT applications, requires adaptive context caching to process and manage enormous amounts of context stemming from IoT. In this work, we propose a novel approach to estimating the context information's demand probability, which helps improve the context retrieval performance of a CMP under near real-time constraints. The proposed approach uses context query logs and applies machine learning algorithms to estimate the context caching probability for context caching. We further use an evolutionary technique for optimising the context caching probability to improve the context retrieval performance of the CMP. We conduct an experimental evaluation using a research prototype CMP, Context-as-a-Service (CoaaS) and show that the proposed technique can significantly improve the context retrieval performance. Analysis of the experimental results showed with context caching probability optimized by evolutionary technique there is an average percentage decrease of 43.68% in the response time of CoaaS.

Context information is in demand more than ever with the rapid increase in the number of context-aware Internet of Things applications developed worldwide. Research in context and context-awareness is being conducted to broaden its applicability in light of many practical and technical challenges. One of the challenges is improving performance when responding to a large number of context queries. Context Management Platforms that infer and deliver context to applications measure this problem using Quality of Service (QoS) parameters. Although caching is a proven way to improve QoS, transiency of context and features such as variability and heterogeneity of context queries pose an additional real-time cost management problem. This paper presents a critical survey of the state-of-the-art in adaptive data caching with the objective of developing a body of knowledge in cost- and performance-efficient adaptive caching strategies. We comprehensively survey a large number of research publications and evaluate, compare, and contrast different techniques, policies, approaches, and schemes in adaptive caching. Our critical analysis is motivated by the focus on adaptively caching context as a core research problem. A formal definition for adaptive context caching is then proposed, followed by identified features and requirements of a well-designed, objective optimal adaptive context caching strategy.

Performance metrics-driven context caching has a profound impact on throughput and response time in distributed context management systems for real-time context queries. This paper proposes a reinforcement learning based approach to adaptively cache context with the objective of minimizing the cost incurred by context management systems in responding to context queries. Our novel algorithms enable context queries and sub-queries to reuse and repurpose cached context in an efficient manner. This approach is distinctive to traditional data caching approaches by three main features. First, we make selective context cache admissions using no prior knowledge of the context, or the context query load. Secondly, we develop and incorporate innovative heuristic models to calculate expected performance of caching an item when making the decisions. Thirdly, our strategy defines a time-aware continuous cache action space. We present two reinforcement learning agents, a value function estimating actor-critic agent and a policy search agent using deep deterministic policy gradient method. The paper also proposes adaptive policies such as eviction and cache memory scaling to complement our objective. Our method is evaluated using a synthetically generated load of context sub-queries and a synthetic data set inspired from real world data and query samples. We further investigate optimal adaptive caching configurations under different settings. This paper presents, compares, and discusses our findings that the proposed selective caching methods reach short- and long-term cost- and performance-efficiency. The paper demonstrates that the proposed methods outperform other modes of context management such as redirector mode, and database mode, and cache all policy by up to 60

Context-awareness in Internet of Things (IoT) applications has significant impact on how IoT data can be processed, stored if needed, reused, and repurposed across multiple IoT applications. Emerging Context Management Platforms (CMP) mediate between context providers and context consumers in order to unify access to context and, provide interoperability that allows cross-domain context querying. This paper proposes an approach to adaptive context caching which enables CMPs to serve context queries from multiple IoT applications. It presents the transient nature of context which is a unique challenge when caching context that requires regular refreshing. The paper proposes two adaptive refreshing strategies based on online-estimated lifetimes (i.e., how long before data is estimated to have changed and refreshing is needed) - reactive and full-coverage. They are evaluated by developing mathematical models and simulations. We further assess the impact of different parameters on context cache performance. The results demonstrate the efficiency of adaptive context caching to minimize operational costs whilst preserving good enough refresh rate and compliance with Service Level Agreements.

IoT hardware and application silos have achieved a significant progress in recent years. Yet, they have been unable to realize the full IoT potential, due to the lack of middleware platforms that provide horizontal IoT context sharing and integration. Context Management Platforms (CMPs) enable interoperability between IoT silos, and they provide Context-as- a-Service. CMPs enable real-time context-awareness but often lack appropriate mechanisms to support the evaluation of their performance. This paper proposes a novel scenario-based approach for generating context queries that can be used to assess the performance of a CMP, specifically the caching performance. The proposed approach allows a user to generate realistic context queries based on simulated scenarios. The proposed model has the ability to capture the real-world scenarios and translate them into a sequence of scenes. An ontology-based approach is used for scene modelling. Once the scenes are modelled, the situations are inferred through contextual reasoning. The proposed model reveals that inclusion of context information in the scene makes it feasible to handle uncertainty of situations as expected in the real-world. Based on the situations, context queries are generated using different query templates. To support a fair evaluation of the caching mechanism of a CMP, queries with different complexities are generated from the proposed model.

Exploiting IoT data collected via embedded sensors in mobile devices and things is useful in context-aware applications. Such context-awareness can also be useful for automated vehicles (AVs) which need to be aware of the surroundings for safety and efficiency reasons. Also, scaling up of Human Activity Recognition (HAR) to Group Activity Recognition (GAR) has attracted significant attention recently. One of the crucial elements of every context-aware system is obtaining context data from context providers (CPs) to be able to recognize a group activity (GA) or a situation among a group. In this paper, we extend our previous framework, called GARSAaaS+ (GARSA-as-a-Service+) to interact with external Context Providers to offer services for mobile Group Activity Recognition and Situation Analysis (GARSA) applications in relation to improving the safety for emerging AVs. Also, context-aware data caching in IoT-enabled applications in order to obtain context quickly is an open challenge - in this work, we propose an approach applying caching rules defined using GroupSense-L, which is a specification language for group activities. We demonstrate our proposed middleware via two scenarios to improve group and individual safety when an automated vehicle is involved. We also demonstrate the feasibility of our model and the expressiveness of our proposed model via a range of scenarios.

We envision robots in cities will be increasingly pervasive - forming new computational nodes, connected to each other and to the Internet, adding to the already proliferating mobile, wearable and fixed ubiquitous computing devices. This paper discusses the notion of cooperation schemes to enable such urban robots to work together with each other, with IoT devices and with humans in different modes, and outlines prototype distributed middleware we are building towards this end.

Human activity recognition using embedded mobile and embedded sensors is becoming increasingly important. Scaling up from individuals to groups, that is, Group Activity Recognition (GAR), has attracted significant attention recently. This article proposes a model and modeling language for GAR called GroupSense-L and a novel distributed middleware called GroupSense for mobile GAR. We implemented and tested GroupSense using smartphone sensors, smartwatch sensors, and embedded sensors in things, where we have a protocol for these different devices to exchange information required for GAR. A range of continuous group activities (from simple to fairly complex) illustrates our approach and demonstrates the feasibility of our model and richness of the proposed specialization. We then conclude with lessons learned for GAR and future work.

With the proliferation of Internet of Things (IoT) devices, there is a demand for technologies to support high-velocity, dynamic resource provisioning to provide secure, cost-efficient, and real-time IoT services in resource-constrained environments. Conventional fog computing by itself cannot address such requirements and needs to be complemented with opportunistic fog computing, by providing mobile fog resources on-demand. In this paper, we discuss key issues in this area, and investigate potential solutions from existing work. We conclude this paper with a summary of gaps, and propose an opportunistic architecture for future work.

Human activity recognition using embedded mobile and embedded sensors is becoming increasingly important. Scaling up from individuals to groups, that is, group activity recognition, has attracted significant attention recently. This paper proposes a model and specification language for group activities called GroupSense-L, and a novel architecture called GARSAaaS (GARSA-as-a-Service) to provide services for mobile Group Activity Recognition and Situation Analysis (or GARSA) applications. We implemented and evaluated GARSAaaS which is an extension of a framework called GroupSense (Abkenar et al., 2016 IEEE 30th International Conference on Advanced Information Networking and Applications (AINA), 2016) where sensor data, collected using smartphone sensors, smartwatch sensors and embedded sensors in things, are aggregated via a protocol for these different devices to share information, as required for GARSA. We illustrate our approach via a scenario for providing services for bush walking leaders and bush walkers in a bushwalking group activity. We demonstrate the feasibility of our model and expressiveness of our proposed model.

Group Activity Recognition (GAR) has attracted widespread interest in recent years. This paper proposes a model and a specification language for group activities called GroupSense-L, and a generic IoT-enabled architecture for integrating individuals sensor data and embedded sensors in ”smart things” to infer Group Activity (GA) based on rule-based reasoning. Also, we propose an incentive model through Group Activity Perspectives and a billing model which benefit both a tracker (one who is interested in getting informed of group activity) and trackee/s (the ones who are being tracked). We demonstrate the feasibility of our model and expressiveness of the proposed language for GAR.

Human activity recognition using embedded mobile and embedded sensors is becoming increasingly important. Scaling up from individuals to groups, that is, Group Activity Recognition (GAR), has attracted significant attention recently. This article proposes a model and modeling language for GAR called GroupSense-L and a novel distributed middleware called GroupSense for mobile GAR. We implemented and tested GroupSense using smartphone sensors, smartwatch sensors, and embedded sensors in things, where we have a protocol for these different devices to exchange information required for GAR. A range of continuous group activities (from simple to fairly complex) illustrates our approach and demonstrates the feasibility of our model and richness of the proposed specialization. We then conclude with lessons learned for GAR and future work.

Numerous types of sensor data can be gathered via devices on mobile sensors, such as smartphones and smartwatches as well as things endowed with sensors. Such sensor data from disparate sources can be aggregated and inferences can be made about the user, the user’s physical activities as well as the physical activities of the group the user is part of. A perspective on this is that the group’s physical activity becomes an explanation for the sensor readings now obtained from this set of sensors. This paper proposes an explanation-based perspective on reasoning about multi-device sensor data, and describes a framework called GroupSense that prototypes this idea.

The population of the world is aging and one of the main concerns of the aged care industry is to provide appropriate care for elderly people as their health and independent functioning declines. This paper reports a systematic review of the roll of Socially Assistive Robots (SAR) in elderly wellbeing, based on Cochrane principles. Relevant publications from diverse databases, including healthcare, engineering, and robotics were sourced and screened. Ninety-five studies in forty-two study groups have been synthesized. The reported outcomes have been categorized based on five PERMA constructs (Positive emotion, Engagement, Relationships, Meaning, and Achievement) in addition to a sixth category (other effects). The findings indicate that SAR has the potential to enhance elderly wellbeing and decrease the workload of caregivers. Based on concerns that emerged during the quality appraisal process, several significant recommendations are made to improve future research and its applicability. Furthermore, acknowledging individuals' needs, expectations, and preferences alongside multi-modal interaction and data collection translates into improvement of personalization of care. The development of new approaches such as web-based interfaces and cloud computing are highly recommended as the means of overcoming the constraints of the limited computing and storage capabilities of SAR. Moreover, human-like engagement characteristics of socially assistive robots should be seamlessly integrated with other assistive technologies such as tele-health, e-health, and smart homes.

Human activity recognition using embedded mobile and embedded sensors is becoming increasingly important. Scaling up from individuals to groups, that is, group activity recognition, has attracted significant attention recently. This paper proposes a model and specification language for group activities called GroupSense-L, and a novel architecture called GARSAaaS (GARSA-as-a-Service) to provide services for mobile Group Activity Recognition and Situation Analysis (GARSA) applications. We implemented and evaluated GARSAaaS which is an extension of a framework called GroupSense where sensor data, collected using smartphone sensors, smartwatch sensors and embedded sensors, are aggregated via a protocol for these different devices to share information, as required for GARSA. We illustrate our approach via a scenario for providing services for tour leaders aiding Vehicle-to-Human (V2H), Vehicle-to-Group (V2G) and Vehicle-to-Vehicle (V2V) interactions to increase the group safety. We demonstrate the feasibility of our model and expressiveness of our proposed model.

Human activity recognition using mobile sensors is becoming increasingly important. Scaling up from individuals to groups, that is, Group Activity Recognition (GAR), has attracted significant attention recently. This paper investigates energy consumption for GAR and proposes a novel distributed middleware called GroupSense for mobile GAR. We implemented and tested GroupSense, which incorporates a protocol for the exchange of information required for GAR. We also investigated the battery drain of continuous activity recognition in a range of simple GAR scenarios. We then conclude with lessons learnt for GAR.