Andrew D. Maxwell

Real-Time Remote Access Laboratory With Distributed and Modular Design

Remote access laboratories (RALs) are online environments for operating instruments and collecting measurement data over the Internet. Such systems are often deployed by universities to support undergraduate students and generally follow the client-server paradigm. This paper discusses a RAL system that enables peer-to-peer (P2P) experimental design and sharing. For this, a modular design is required, which allows participating nodes to create rigs and host those individually at distributed locations. The proposed architecture is generic and can be used with any distributed P2P network control systems over the Internet. In this paper, a distributed remote control framework is presented with regard to a P2P RAL system. The experiments in the RAL require three subsystems handling the user interface, instruction interpretation, and instruction execution which can be organized and operated in different manners depending upon the experiment. The key component for creating and controlling experiments is the microcontrollers that can be easily obtained, configured, and set up for use over the Internet. The most popular microcontrollers are examined for suitability to the distributed control architecture. The basic layout of a message-based network protocol suitable for programming the devices and communication between peers for remote instrumentation and control is discussed, and queuing and flow control mechanisms are compared and tested for the proposed framework.

Robot RAL-ly international - Promoting STEM in elementary school across international boundaries using remote access technology

Engaging school children early in STEM activities plays an important role in their choice to study engineering in later years. This paper describes a pilot project where Remote Access Laboratory technology at a university is employed in an inquiry-based learning activity with elementary school children in Japan and Australia. Investigation into how RAL technology facilitated collaborative learning in the K-12 demographic was then examined. Children in Japan designed a track that was constructed by their peers in Australia. The Japanese students then navigated the track in Australia with remotely controlled robots using the RAL system. A number of camera feeds allowed the students to observe the robots and the track. Both groups of students, as well as the participant researchers, took part in co-constructed focus group discussions after the event. A thematic analysis indicates that these activities provide students with opportunities for rich learning experiences in science, math and technology. Engaging young children in STEM activities provides a strong pathway to a better understanding of science concepts and ultimately a career in engineering.

An overview of system architectures for Remote Laboratories

Remote Access Laboratories have been successfully installed around the world enabling students to practice practical skills and develop their knowledge through experimentation. Remote laboratories have the advantage of being accessible from anywhere anytime allowing users more flexibility and mobility. Typically installations are designed to resolve issues of access or utilization. Clear pedagogical design is now seen as a critical development in the instigation of new experiments, and access systems. This paper examines existing system designs, using a Program Logic analysis to identify key inputs, outputs, and impacts for the creation of systems. It aims to identify commonalities, or differentiation with respect to the original goals. Through this analysis it was identified that many RAL systems implement direct mimic user interface design focusing on individual access to a kinesthetic learning experience. Learning design is also typically implemented by the system and experiment designers, potentially reducing the impact of the system or activity usability.

Using network enabled microcontrollers in experiments for a distributed remote laboratory

Remote access laboratories are being used by universities to provide access to experiments through web based online environments. Usually, such experiments setups are an extension of the real hardware used in on-site laboratories. For a maker oriented RAL, users are encouraged to build their own experiment setups based on their understanding of the concepts and share them with their peers. Such a RAL system is targeted towards school level experiments. This poses a new problem of compiling equipment from an improvised set of tools, typically available at home or in schools. This is very different from the sophisticated laboratory equipment used in universities. In this paper, characteristics of such a RAL system are discussed and commercially available microcontroller boards to implement them are compared. Methods to operate these rigs based on a messaging protocol and an analysis of their performance in communicating with the network are presented.

Joining the game and the experiment in peer-to-peer remote laboratories for STEM education

Remote Access Laboratories (RAL) provide access to experimental setups from remote locations. These experimental setups are composed of controller units programmed to gather data and interact with user inputs. A distributed version of RAL can be maker oriented i.e. the experiment rigs are designed by individuals and shared among each other. This paper presents the programming aspects and activity user interface (UI) design and organization of experiments in a distributed RAL aims at STEM education. The user interface must be interactive to increase engagement and motivation for the user. Being designed for school students, the environment to create the control logic of a rig created by the student needs to be on a homogenous platform. The programming language has to be easy to understand and use. Characteristics and requirements of a graphical programming language SNAP, which is modified and used as the programming platform for RAL, is studied in this paper.

Estimation of round trip time in distributed real time system architectures

Real-time distributed systems often utilize centralized nodes to manage and coordinate the exchange of information; in particular, for users to are unable to communicate directly because of their location behind firewalls or proxy servers. Identifying these nodes is crucial for successful operation of the entire distributed system. The geo-locational placement of these relay node can be determined based on network properties such as Round Trip Time (RTT). In this paper a real-time e-learning system - Remote Access Laboratories (RAL) are discussed as an example of a distributed real-time system. RTT is an important factor directly impacting the Quality of Experience (QoE) of users. In a distributed RAL environment, there may be multiple users who can be both learners (service users) and providers in an activity. As such, the equipment and corresponding learning materials are not concentrated at one location, but spread over multiple nodes. A method to determine the suitable node or relay based on population distribution in a geographic region and projected RTT based on distance between them is proposed here. A mathematical model and simulation data are presented to support the effectiveness of the method. The results are compared with peer-to-peer network model, which is optimal, to establish the difference in performance.

Performance and Quality of Experience of Remote Access Laboratories

Remote Access Laboratories have become important learning and teaching tools. This research presents a performance study that targets a specific remote-access architecture in a university environment. The system provides authenticated and mediated global remote access to virtual machines as well as hardware systems, which drive physical experiments. This paper presents system performance results that have been obtained with a set of automated and user-based tests. Key aims of the study were: To gain a better understanding of traffic that is caused by experiment usage; to get an indication of expected user performance; and to develop a measure to predict Quality of Experience, based on easily measurable Quality of Service parameters. The study emulates access bandwidth, round trip times of typical usage scenarios and provides results that allow classifying expected user performance. It demonstrates that failure rate is excellent measure of usability. Thin-client and remote desktop architectures are popular to separate the location of users and the actual data processing and use similar structures. Results of this study can be applied to these applications as well.

Performance evaluation of network architectures for collaborative real-time learning systems

Online learning tools are commonly used for teaching distance education students in tertiary education. These tools rely on Internet connectivity, and network performance impacts on how these systems are used. Network performance is particularly relevant for interactive, real-time interactions such as video conferencing, online control and interactive games. Systems that host these learning systems are generally hosted in a central location, often where the institution is situated. This paper presents a study that compares the performance of such a central system with a system that uses peer-to-peer connectivity for real-time interactions between students. Using Australia as an example, the two architectures are simulated. Numerical results demonstrate the advantages of the peer to peer architecture in this context.

A framework for analyzing and evaluating architectures and control strategies in distributed remote laboratories

Remote Access Laboratories (RALs) have been used to develop experimental knowledge about practical engineering topics for a while. Distributed remote laboratories aim to share experiment among institutions and individuals through a distributed architecture. Experiments from diverse areas are combined as part of a larger system. Multiple control strategies are used to integrate experiments in Remote Laboratory Management Systems (RLMSs). This work defines two main categories to analyze the various implementations, white box and black box approaches. Experiments can be on a spectrum between these two extremes, sharing properties of both. When integrating an existing experiment into a new distributed RAL system, it is useful to evaluate the experiment with respect to its host or new RLMS for determining the best strategies to assimilate it. This paper provides a framework for such evaluation based on a number of properties of experiments. The proposed framework is called SHASS (Software, Hardware, Assessment, Support, and Share-ability) based on several factors such as the hardware used, the software to create the program, methods of sharing, users support, and assessment of users performance. It can be used to evaluate quality and identify options for improvements within an experiments existing RLMS as well. Using this framework, a black box and white box approach are compared using two examples - federated and Peer-to-Peer RAL. The evaluation focuses on technical capabilities and development possibilities. A set of four experiments are also analysed to illustrate the utility of the framework in creating and improving experiments with respect to their RLMS.

Components relationship analysis in distributed remote laboratory apparatus with data clustering

Remote Laboratories are network controlled systems operated by human users through the internet for educational purposes. A distributed version of the remote laboratory requires the experimental rigs to be designed by individuals thus making it difficult to obtain formal models of the experimental rigs. A rig consists of a micro-controller unit with multiple ports to connect sensors and actuators. This paper proposes a timed automaton based model of experimental rigs that can be common to all sites. Further, the relationship of components of a rig is analyzed based upon this automaton. The components can be grouped into multiple sets where each set has two properties - the bond between each component in the rig and how frequently they are accessed. A method to obtain the component sets and to determine these two characteristics using data clustering is described.