Ac 2011-376: Wireless-integrated Embedded Real-time Con- Trol: a Case Study in Adopting Resources for Develop- Ment of a Low-cost Interdisciplinary Laboratory Project
چکیده
In the last decade, it has become apparent that the grand challenge problems of this century span disciplines. In spite of this, engineering curricula are still strongly stovepiped, even within each engineering discipline, and both inertia and downward budget pressures encourage curricular conservatism. At the same time, the need is urgent to expose students to the diversity and complexity of real-world problems where there is no “best” solution. How should we help students learn across disciplines and blend disciplinary knowledge to solve problems? This paper describes a laboratory project suitable for courses in areas of control and embedded systems that weaves critical aspects of control systems design with real-time embedded systems hardware and software, and along the way incorporates additional skills and tools. The project builds on previous efforts that have used the classic “ball-in-tube” experimental platform. We have developed an extremely low-cost experimental platform that student teams assemble from simple parts (e.g., shoeboxes and muffin fans), and that uses wireless communication between the real-time platform and a personal computer that provides a human interface and analytical tools. For real-time data acquisition and control, we adopted the CLIO platform that was designed for the experiential component of MUSE (Multi-University Systems Education, www.uvm.edu/~muse), an NSF-sponsored pedagogical effort to increase the ability of students to become conversant in skills related to systems thinking. In this spirit, the work discussed herein exposes students to experimentation, modeling and design across system layers. While tackling the project, students have also become more adept at (i) architecting distributed applications that integrate embedded and desktop computing systems, (ii) data acquisition, including measurement noise and signal conditioning, (iii) actuation, including motor control, and (iv) wireless communication. We present early assessment results evaluating how effectively the project helps students build critical systems-thinking skills, and the challenges of adopting resources for fast-tracking the development of new laboratory projects. Introduction The notion of systems thinking is well-known, but views vary on its specific definition. For example, the systems dynamics community emphasizes understanding the temporal dynamics of interconnected parts, including the effects of feedback and emergent behaviors, via conceptual and simulation models. Others have emphasized design in a broader context, including assessment of societal impacts and awareness of economic and societal goals. In our practice-oriented view, systems thinking is already pervasive across engineering disciplines, highly valued in industry, and normally learned on the job. The best systems thinkers become technical leads and managers in part because they become experts in systems thinking as part of their professional and technical career growth. Systems thinking is in our experience often developed through informal mentoring, and is larger in its aims and scope than systems engineering (as often executed, e.g., using trade studies involving several variables). Our definition includes the systems dynamics viewpoint and encompassing contexts, but also involves thinking about the design process in ways that (i) span traditional disciplinary boundaries, (ii) integrate design layers from high-level abstractions to low-level implementation, and (iii) reveal the relationships that couple seemingly disconnected models. Two aspects are critical: modeling and context; context may be broad, including human, societal, and environmental impacts. Our interest in systems thinking is driven in part by our motivation to help engineering students develop the skills that will be critical to the pressing technological challenges of this century. While it is widely recognized that the engineering graduates of today and the future should have better systems thinking skills, the challenge of inculcating those skills remains. In this paper, we describe an attempt to do this within the constraints of an embedded systems course that is typically found in an ECE curriculum. Overview of the Approach The focus of this effort was the development of a final laboratory project in the context of a senior-level EE course. The course, EE 410 – Embedded Control, focuses on deeply embedded systems, and has been offered by one of the authors four times in the last decade at Northern Arizona University, with a range of topics including robotics. With each iteration, the course content has been updated to keep up with industry trends, including the transition from assembly language dominance to the use of the C language, moving from 8to 16-bit architectures, the evolution of multiple clock sources/domains and low-power modes, and the increasing importance of networked embedded systems, especially implemented using wireless communication. The course strongly emphasizes experiential content: the laboratory projects account for 80% of the final grade. Student teams consisted of at most two students. For all projects, a team’s project grade is based on both its degree of success in demonstrating achievement of objectives in the laboratory and a comprehensive written report. To help students equip themselves with the skills to tackle a substantive final project that involves systems thinking, we adopted a progressive learning method, with continual reinforcement and synthesis of skills and tools via the sequence of projects (Table 1). Table 1: Course projects leading up to Final Project. Project Concepts Skills and Tools 0: CLIO Quick Start Digital I/O review; looping and software delays; program compilation and loading; port/pin configuration Integrated development environments; familiarity with hardware platform and MSP430 MCU; hardware interfacing 1: Heartbeat Simulation Functions; parameterized software design; real time systems Flowcharts; defined constants 2: Switch-controlled Heartbeat Simulation Interrupts and program state; externally-triggered interrupts; digital inputs; mechanical switch characteristics Interrupt service routines (ISRs); switch debouncing 3: Software LED Dimmer with Switch Control Clock and clock domains; timers; pulse-width modulation (PWM); internally-triggered interrupts Software integration of multiple functions; timer modes; timer configuration and management Table 1 (continued): Course projects leading up to Final Project. Project Concepts Skills and Tools 4: Pseudorandom Number Generator and Reporting to PC Linear feedback shift registers; maximal-length sequences; random and pseudo-random processes; statistics: sample mean and autocorrelation; user interfaces Advanced bit-oriented processing in C; serial communication; UART programming 5: Temperature Sensing Sensing; analog-to-digital conversion (ADC); periodic schedulers; MCU low-power modes ADC peripheral configuration and operation; low-power modes on the MSP430 via ISR’s; PC communication (more) 6: Motor Control Actuation; DC motors; motor control; power electronics; pulsewidth modulation H-bridge interfacing with MCU; hardware-based PWM 7: Transducer Characterization Data acquisition; transducer signal processing; resolution; noise, bias, and non-linearity Advanced ADC peripheral configuration and operation; data conversion The linear project sequence was designed to cover a cloud of concepts that span from the general—traditional ECE subdisciplinary silos (Figure 1, bottom)—to specific conceptual components required for the final project. Thus the concept map of the course topics and how the development of concepts leads to the final project is hierarchical in nature, and the final project integrates knowledge from across subdisciplines as shown in Figure 1. Figure 1: Hierarchical concept map. Connectedness of concepts is shown by overlapping nodes in addition to edges. Note that all possible links and overlaps are not shown.