Lectures: Monday, Wednesday, 2:30 - 3:50, Hamburg Hall 2224

Wearable Computer Laboratory, Room 2201 Hamburg Hall, 8-6465

Wearable Software Laboratory, Room 2205 Hamburg Hall, 8-3267

Instructors: Dan Siewiorek 1201 Hamburg Hall, Ext. 8-5228,dps@cs.cmu.edu

Asim Smailagic 1217 Hamburg Hall, Ext. 8-7863, asim@edrc.cmu.edu

Jim Garrett 123A Porter Hall, Ext. 8-5674, garrett@cmu.edu

Jane Siegel 1303 Wean Hall, Ext. 8-6764, jane.siegel@cs.cmu.edu

John Stivoric 2203 Hamburg Hall, Ext. 8-7890, js1y+@andrew

Brian Gollum 2205 Hamburg Hall, 8-8401, brig+@cmu.edu

Secretary: Mrs. Laura Forsyth, 4125 Wean Hall, Ext. 8-2619 or 8-2570

Electronic mail address - forsyth@ cs.cmu.edu

Office hours: 8:30 am to 5:00 PM

Paradigm Shift in Computing

The information processing industry is undergoing a paradigm shift. Commencing in 1960 information processing was concentrated in mainframe computers operated by central staff and accessed by custom-built programs executed in batch mode. By 1970 the invention of the time-sharing operating system allowed users to interact with their information on-line. However, time-sharing systems were still centrally based with most of the computing cycles devoted to information manipulation rather than the human computer interface. With the advent of the personal computer in the early 1980's a substantial portion of the computing power could be dedicated to the single user. New paradigms such as the spread sheet allowed the user to interact with their data on an item-by-item basis looking for patterns and playing "what if" scenarios. Since 1980 technology has been devoted to shrinking the size and weight of personal computers without substantially changing the way users interact with their computing environment. Conventional input/output devices place an ultimate limit on the size and weight of personal computers. Size is limited by the conventional typewriter-like keyboard whose dimensions have not changed substantially for over one hundred years. Both size and weight are limited by displays the size of notebook paper intended to be viewed from several feet. The size of the display places a lower bound on the personal computer's energy consumption and hence weight primarily dictated by the weight of the energy storage devices such as batteries.

The convergence of a variety of technologies makes possible a paradigm shift in information processing for the 1990's. Continued advances in semiconductor technology makes possible high performance microprocessor requiring less power and less space. Decades of research in computer science have provided the technology for hands-off computing using speech and gesturing for input. Miniature heads-up displays weighing less than a few ounces have been recently introduced. Combined with mobile communication technology, it is possible for users to access information anywhere. It is indeed possible to sense a user's position so that the information can be superimposed upon the user's workspace. Mobile computing deal in information rather than programs, becoming tools in the user's environment much like a pencil or a reference book.

Sensors make the computing system an active part of the environment. When the user modifies the environment, e.g., by terminating a cable, the information could be automatically entered into the system. The lack of accurate, timely updating mechanisms for information about physical systems is one of the key limitations to the use of on-line databases. The mobile computer provides automatic, portable access to information. Furthermore, the information can be automatically accumulated by the system as the user interacts with and modifies the environment thereby eliminating the costly and error-prone process of information acquisition. Much like personal computers allow the accountants and bookkeepers to merge their information space with their workspace (i.e., a sheet of paper) mobile computers will allow mobile processing and the superposition of information on the user's workspace.

This is a project-oriented course which will deal with all four aspects of project development: the application, the artifact, the computer-aided design environment, and the physical prototyping facilities. The class, in conjunction with the instructors, will develop specifications for a mobile computer to assist in inspection and maintenance. The application will be partitioned between human computer interaction, electronics, industrial design, mechanical, and software components. The class will be divided into groups to specify, design, and implement the various subsystems. The goal is to produce a working hardware/software prototype of the system and to evaluate the user acceptability of the system. We will also monitor our progress in the design process by capturing our design escapes (errors) with Orthogonal Defect Classification (ODC). Upon completion of this course the student will be able to: generate systems specifications from a perceived need; partition functionality between hardware and software; produce interface specifications for a system composed of numerous subsystems; use computer-aided design tools; fabricate, integrate, and debug a hardware/software system; and evaluate the system in the context of an end user application.

The application will be either improving the efficiency of off-shore oil platform crews or of geographically distributed design teams.

In selected cases, off-shore oil platform crews are being downsized from 30 to 6 personnel. Crew will be cross-trained and be more generalists than the current practice of specialist training. The crew will depend more on electronic technical manuals and assistance from personnel on-shore and on other platforms to solve maintenance and operation problems. 

The Infocator is a small yet high performance computer with wireless connectivity for exchanging information between groups of cooperating workers. There are many examples of time volatile information whose usefulness is overcome by events even before it is committed to the world wide web. How can the right person provide the right information at the right time without becoming more of a burden than e-mail, telephones, and pagers?

We are currently working with clients to determine which application will be the focus for the class.

The course is divided into four major phases (Conceptualization, Planning, Design, and Implementation), each composed of up to several sub phases.

• Problem Definition. The goal of this sub phase is to define the problem which is being solved, perform requirements analysis, and evaluate user needs. A variety of brain-storming techniques will be employed to develop a product design definition including attributes such as functionality, cost, performance, technology acquisition, and fabrication techniques.
• Technology Survey. The final shape of a system is often determined by what technology is currently available. A survey of available technology, with special emphasis on light weight displays and input devices, will further refine the Product Design Specification. Lessons learned from prior generations of wearable computers will be discussed. New components such as spread spectrum radio and VGA displays will be acquired and interfaced to existing wearable systems to determine the feasibility and complexity of the new technology. Video tapes of current practice as well as discussions with end users will generate interactive scenarios.

• System Architecture Specification. Given the constraints of available technology and the user's computational environment, the architecture for the system will be developed. Topics such as local versus distributed processing, position sensing, computer/human interface, and information updating must be addressed by the selected architecture. Planning will also include interdependencies between the technologies, people, and resources available in the course.
• Subsystem Specification. The system functionality will be partitioned and assigned to hardware and/or software components. Refinement of the Product Design Specification will include attributes of the subsystems including performance, interface, and evaluation criteria.

• Detailed Design. Each subsystem will be designed and implemented. Design will be supported by contemporary computer-aided design tools. Mini workshops as necessary on relevant tools will provide students a basic introduction to the state of computer-aided design. Human computer interaction studies will be designed in conjunction with mechanical/electronic/software mock-ups to provide data for design decisions.

• Implementation. The detailed hardware/software designs will be implemented using both on-campus and off-campus facilities. On-campus physical prototyping facilities will be used when appropriate. The state-of-the-art in rapid prototyping will be presented.

• System Integration. The various hardware and software subsystems must be individually tested and then integrated into a working system. System integration and testing plans will be formulated commencing with the system architecture specification phase. The system will be evaluated through controlled user experiments.

• Methodology Evaluation. As a final phase, the methodology followed in the course will be quantitatively and qualitatively evaluated and modifications suggested.

System design is a science best acquired through experience. The course is project oriented with a series of phases. In general, each phase will culminate with a written design document and an oral design review. Later phases will also include a demonstration. Students will be divided into project teams whose performance will be graded in terms of classroom participation, written documents, oral presentations, design methodology forms, and demonstrations. Individual feedback will be given in performance reviews at the end of each major presentation. The relative weight for course grading for each of the phases is roughly as follows:

• Technology Feasibility, System/Subsystem Specification - 20%
• Detailed Design - 30%
• Implementation - 30%
• Integration - 20%

Reports Format 

In order to capture the relationship between the evolving portions of the design and design process, information will be entered into an on-line repository formed by Web pages. Hyper-links will be added between critical decisions. As the design evolves, changes that impact other design decisions will be easier to identify. The Web pages will also allow students joining the project later to review the design history. The goal is to provide templates and information that can be used in future courses to improve productivity and minimize design escapes. The report for each phase will follow a standard format and include:

Product

• Product Design Specification
• Attribute Dependency Chart for design parameters
• Design Alternative Evaluation Chart

Design Process

• Inter- and Intra-group dependencies
• Cumulative time line (actual and projected)
• Personnel assignments
• Person-hours effort as a function of task (actual and projected)

Reports will build upon each other and be successive refinements of previous reports. The goal is to document the design evolution as well as the design process.

The class will revolve around project groups. Each group will have the responsibility of designing and implementing one of the major subsystems as well as interacting with other groups to ensure compatibility. Groups will consist of from two to five students, depending on the complexity of the subsystem. A project management council will meet weekly to discuss logistical issues. The council will be composed of the course instructors and a rotating member from each project group. Liaisons between groups will also be utilized. Potential project groups include:

• Hardware. Survey available hardware. Determine what should be purchased and what should be custom designed. Interface with other groups to minimize modifications to standard software due to hardware design decisions. Purchase components and external manufacturing services.
• Mechanical. Generate the housing to support the hardware and selected human interface. Purchase components and external manufacturing services.
• Software. Select, port, and adapt operating system software to unique features of the hardware which can include: new device drivers for the display, input, and secondary storage; new features such as power management, etc. Produce the application data and software.
• HCI. Determine the human-computer interface. Perform the requirements analysis. Design and conduct intermediate experiments to support design decisions. Devise the hardware and software for determining the user's relationship to the displayed information. Conduct a final evaluation of the product and the design methodology.

The major form of communications outside of class will be electronic. In addition to e-mail each group will have a bboard as well as bboards for class announcements and class wide discussions:

academic.cit.39-648.sw

academic.cit.39-648.me

academic.cit.39-648.el

academic.cit.39-648.hci

academic.cit.39-648.announce

academic.cit.39-648.discuss

The class also has a home page:

/afs/cs.cmu.edu/`wearable/class/spring99