Michael V. Doran

A cognitive-based approach to introductory computer science courses: lesson learned

A project has been undertaken this past year using a cognitive based approach to implement the Computing I and II courses as defined in Implementation D of Computing Curricula 1991. The salient features of this project include: (i) strategic sequencing and associated levels of mastery of key topics based on Bloom levels; (ii) a spiral approach to presentation; (iii) integral use of structured closed labs; (iv) frequent feedback and (v) early use of teams. This project has led to a series of course documents that explicitly define and schedule course micro-objectives, that map each micro-objective to a specific Bloom knowledge level, and that help to achieve and measure those objectives.

Integrating collaborative problem solving throughout the curriculum

Our graduates are ill-prepared for entry positions in industry. This is the message from a 1994 NSF task force comprised of members from academia and industry. Among the specific deficiencies cited were problem-solving skills and the ability to work in groups. In a recent publication, the authors described a group problem-solving model, Group Zig Zag, based on the Myers-Briggs Type Indicator. In this paper we show how the integration of a collaborative approach throughout the curriculum can be achieved by adopting the Group Zig Zag model.

Improved learning by use of a video knowledge-base

Increased mastery of basic computing theory in the introductory computer science class is the subject of on-going research at the University of South Alabama. Previous and continuing projects have considered tools, labs and complexity of assignments. The focus of this study is the role of video media presentations and rehearsal. The role and benefit of video in the course will be discussed; strategies for the evaluation of the effectiveness will be proposed.

Data: data to algorithm translator analysis

During the past several years, software engineers have developed various programming environments to aid in the software development process. Some of these environments, serving as pedagogical tools have focused on the coding phase (Chandhok et al 1985), (Johnson and Soloway 1983) and (Shapiro et al 1981). It is a firm belief that the emphasis, especially with novice programmers, should be placed upon the design phase. The tools developed should aid in the step-wise decomposition of the problem. This view has been stated by researchers (Dijkstra 1968), (Soloway 1986) and (Wirth 1971). Doran (1986, 1987, 1988) developed an environment based upon a graphical tree representation of an algorithm. This tree, developed and refined by Law (1986), is called a structure chart. These structure charts were used to allow introductory level students to design algorithms for their programming problems. The environment, SLAW, was created to allow students to create these charts and also allows for automatic translation into various programming languages. SLAW has been used in classroom settings in an attempt to allow students who know one language to acquire the syntax of others quickly and easily (Doran 1986). Doran (1988) also developed a prototype system to add intelligent tutoring powers to the environment.

Structured programming environments (abstract only)

Structure charts are a graphic algorithm design tool. An on-line editor has been implemented that can be used to create and maintain structure charts. In addition to the graphic nature of structure charts, a data dictionary is necessary to more completely define the design. A current project in this area involves incorporating the data dictionary into the on-line environment. This additional feature enables automatic coding in a high-level programming language.

Enhancing Robotic Experiences throughout the Computing Curriculum

The study of robotics is often an excellent recruitment motivation for students entering a computing curriculum. It is also often cited that Artificial Intelligence (AI) and related robotics is a critical area of future innovation in computing. However, there are a number of challenges to implementing a robotic curriculum at the university level. A few of those challenges include, lack of a uniform hardware platform, incongruity in the software used, and the missing significant deliverable artifacts by students at the end of a term to encourage learning and further interest. Educational robotic activities are often treated as advanced research topics with high associated costs that prohibit widespread integration into the curriculum. In this paper, we present the deployment of a multi-stage robotic platform that attempts to address these challenges and overcome the obstacles. We detail the evolution of the hardware from using a readily available LEGO Mindstorm base platform to next include the addition of an Arduino and Raspberry Pi. Likewise, we detail the transition from the basic LEGO Mindstorm software to the use of leJOS and PYTHON. At each stage, additional sensors and software libraries can likewise be added. Finally, we present an achievable academic goal for students to use the hardware and software platforms to develop a model autonomous vehicle with aspects of intelligent control, learning and adaptive behaviour. All levels of projects can be undertaken at the appropriate level of maturity for the students in the computing curriculum.

Cybersecurity issues in robotics

Cybersecurity is not highly prioritized during the design and manufacture of robots. As with other embedded systems a higher priority is placed on development costs and delivering functionality to consumers. In the future greater attention to cybersecurity will need to be given as the use of robots continues to grow in the manufacturing, military, medical, eldercare and the automated vehicle markets. This work identifies current and potential cyber threats to robotics at the hardware, firmware/OS, and application levels. Attack scenarios at each level are presented and discussed. Additionally, the economic and human safety impact of a cyber attack on robots is examined. Finally, possible countermeasures are suggested.

Information science (abstract and references only): managing the software lifecycle

Recent research at USA in the DATA (Data to Algorithm Translation Analysis) project has centered on the algorithmic development process [Doran and Longenecker 1989]. In this research, tools and methods are investigated which assist novice students define and monitor the development of their problem solutions. This approach is adequate for an initial exposure to formal analytic problem solving, but it is recognized that as the level of sophistication of the student increases, so must the tools and methodology. Software engineering presents numerous methods and approaches to providing this lifecycle coverage. One of specific interest is that of the software team. The team concept was recently outlined and discussed by Rettig [Rettig 1990]. However, the proper construction and management of teams is often a difficult and misunderstood process. A central idea of information science is the organization and management of the process in a systematic fashion. Therefore, it is our hypothesis, that to better accomplish the goals of information science, software engineering fundamentals must be managed and tools provided. As mentioned above, design tools have already been and continued to be addressed by the DATA project. The focus of this research is the application of information science theory to provide management for more sophisticated software projects, especially in the area of software teams. Previous researchers have addressed both the management and measurement of software development by teams [Henry and Kafura 1981], [Henry 1983] and [Selig and Henry 1988]. We propose the definition and inclusion of a similar tool into the DATA environment. This inclusion provides several benefits: (i) it recognizes that development does not occur in isolation, (ii) team behavior can be monitored and managed, and (iii) a measurement of achievement can be provided.

Problem analysis and solution development mechanism for DATA (abstract and references only): data to algorithm translator analysis

DATA is an educational programming environment currently under development at the University of South Alabama. DATA has evolved from SLAW [Doran and Longenecker 1989], which uses structure charts [Law 1986] to represent algorithms in graphical trees. The main goal of DATA is to teach students to use data flow diagram analysis to solve formal analytic problems. A graphical design language based on structure charting is developed for the tool to represent algorithm and data structure designs. A visual execution system will be developed to display the execution of algorithms and the manipulation of data graphically. This part of the research is to develop a problem analysis and software design methodology for DATA. This methodology is based upon Box Structured Methodology [Mills et al 1987]. The Box Structured Methodology analyses a problem and represents its solution in terms of black boxes (external abstract views), state boxes (internal state views) and clear boxes (internal data abstractions). This research will incorporate a sub-system decomposition and control flow decision mechanisms to the methodology for its state migration and clear box expansion processes respectively. The data flow diagrams developed are transformed to an object oriented design [Alabiso 1988]. The design of control flow among objects is determined by the Law of Demeter [Lieberherr et al 1988]. Currently, a grammar for this mechanism is under development to incorporate with the grammar used in Box Structured Methodology. Future development will automate the methodology by implementing it into DATA.

Increased productivity using a preprocessor for Dataflex Fourth Generation Database Language (abstract)

Reusable code and higher generation languages have been sited as methods for systems development which will cause reduction of time to develop new code, and which will lead to significant reduction of developmental errors. In each generation of language syntactic elements of the language decompose into multiple statements at a lower level. Dataflex is a 4th GL associated with a hierarchical database. While it is an efficient language, it nonetheless consumes considerable numbers of statements to solve routine problems not solvable with its standard macros. Many routines seem to be repeated during development of many similar systems. A technique common among professional programmers is to clone whole programs. We proposed to build a preprocessor to automate the cloning process. It would either recognize keywords in a statement and then use associated parameters in an expansion process, or if no keywords were present, the statement would be copies without expansion. A scrolling window display routine command sequence was developed to test the preprocessor. Productivity gains of 25-50 to 1 were obtained with reduction or elimination of errors. While scrolling window routines are not inherently difficult, as much as 500 lines of code are required to handle a simple window, including editable fields, lookup or display-only type fields. To implement such a window #lines = 3 + # of fields in a line using the preprocessor. In development of accounting systems, posting routines were cloned wherein similar fields in one file were posted to fields in another file. Copy and update preprocessing macros increased coding efficiency by at least 25 to 1 these routines.