Brainstorm: The Most and Least Useful Courses for Engineers
Q: What specific courses were the most or least useful to you for your engineering career?
By David Paloian, academic program manager, Microchip Technology
The purely math courses were the least useful when I was taking them and early in my career.
If more time was taken to help me think mathematically (rather than mechanically doing the math) and to know when and how to apply certain mathematical concepts and techniques to particular problems, I would have found the math more useful from the get go.
Cliff Ortmeyer
global head of solutions development, Newark element14
By Cliff Ortmeyer, global head of solutions development, Newark element14
Enrollment in undergraduate engineering programs in 2017 was the highest it has been in the last 10 years, according to the American Society for Engineering Education. The courses in these programs cover the core elements used in almost all types of circuit design and provide a strong foundation for general engineering knowledge.
The most helpful courses to me were:Throughout your career, you will continue to be exposed to and learn a large range of technologies. Additionally, as most applications have hardware and software elements to them, having a broad hardware background allows you to have a greater system understanding. We all have to learn to program at some point, and those skills will continually evolve, but having a strong analog and digital circuit understanding will also help to demystify new technologies as they emerge.
- Electronics 101 and 102: These are the first classes that all electrical engineers take. The fundamentals stick with you throughout your career, no matter what area you go into. From Ohm’s law to understanding the essentials of how active and passive components work in both the time and frequency domain, these are the core building blocks most everything else is built upon.
- Electromagnetics: A much higher level course, but again the core concepts are used in a multitude of applications. In the early days of my career, this class gave me the background to understand and figure out how to design and measure noise generation and mitigation—whether it was radiated or conducted. Now with the continued emergence of IoT and a wirelessly connected world, understanding the basics of electromagnetics helps to simplify and put into context the various standards and protocols, the choice of frequencies, and why they are used in various applications.
- Fundamentals of analog electronics as well as master’s level power electronics courses:Even in these days of modular/digital design, somewhere along the way you are most likely controlling an analog signal. Everything from how that signal is affected and manipulated to why it overheats and fails, all boils down to how you are controlling that current. As processing speed continues to increase and larger amounts of rapidly drawn current is needed (in the most efficient manner, let’s not forget), a firm understanding of the “analog” world is still a necessity.
Harsha Nanduri
sr. marketing engineer, Microchip’s Development Systems BU
By Harsha Nanduri, sr. marketing engineer, Microchip’s Development Systems BU
One of the courses that made me comfortable working with mechanical, electronic, and software systems was my Control Systems Engineering course.
The course covered in detail core concepts related to optimization of systems and analyzing its properties.
This course helped me understand electrical engineering at a system level and understand some of the engineering design choices made and why they are made.
Alessandro Piovaccari
chief technology officer, Silicon Labs
By Alessandro Piovaccari, chief technology officer, Silicon Labs
Since the mid-1960s, IC design has required a wide range of know-how, from underlying physics, manufacturing process technology, and device-level layout to transistor-level and logic design, including coding software to solve KVI equations or perform logic minimization. Electrical engineering jobs required Renaissance scientists, knowledgeable and agile in many disciplines. Engineering students had to take many foundational classes, including advanced calculus and advanced physics, requiring a deep understanding of electromagnetics and programming, all the way to the transient analysis of binary bits. Bachelor’s degrees were almost generic science degrees, while all specialized education happened at the Master’s degree level and above.
Since the early 1980s, three decades of VLSI have exploited the fulfillment of Moore’s Law, the success of CMOS technology, and the development of sophisticated CAD tools and methodologies, culminating in the system on chip (SoC) era, which fueled the PC, internet, and wireless revolutions. The increasing complexity of SoCs and the significant development investments needed for tools and engineering know-how required the creation of large teams of very specialized engineers. This was made possible by a hierarchical approach, where product development was split into independent phases, built on top of each other, where the lower level and earlier phases could be abstracted at a “black box” level. This hierarchical approach has proven to be very powerful, enabling exponential progress in complex product development.
All of the electronic devices we are using today, including computers, smartphones, tablets, and modern cars are a tangible proof of the success of this hierarchical engineering approach. To be more specific, engineers working at a higher application level (e.g., software developers) did not have to know the details of how hardware and software components were designed (e.g., knowledge of IC design) or even how they were manufactured (e.g., built-in semiconductor foundries). As a consequence, engineering studies became increasingly specialized. Electrical engineering departments were split into multiple departments such as computer science, telecommunications, and so on. Around the world, more advanced, specialized courses started to be taught at the undergraduate level at the expense of dropping fundamental, wider range courses such as advanced physics, electromagnetic fields, and nonlinear systems. Did an engineer whose only job was to conduct static timing analysis need to know about Maxwell equations and radio signal fading phenomena? Did an RF mixer designer need to know about object-oriented programming?
Now, at the apex of this era, manufacturing, tools, and methodology maturity, and the wide availability of specialized talent make SoC design problems less challenging. The amount of complex building blocks available to any hardware or software developer has reached a level of sophistication that was not even imaginable just 10 years ago. Complex microcontrollers that have more power than a personal computer of the early 1990s can be purchased for just few dollars, and they can operate for years on batteries.
One could argue that even more specialization is still possible, or maybe even required. But the reality is, we live in a world of more or less linearly increasing resources (some could argue limited, but I think the limit is still far away), while, with the current approach, the exploding IoT “re-evolution” will require an exponential growth of engineering resources. For instance, we cannot afford to increase tenfold the current production of batteries to power these devices, both in ecological and economical terms. We not only need to improve the battery quality, but also reduce the power consumption of connected devices. In addition, a massive deployment of IoT devices requires a strategy of continuous cost reduction that cannot be fueled alone by an already compressing Moore’s Law, due to a physical world that cannot be infinitely scalable (at least until we achieve the promise of quantum computing). Finally, the complexity of these IoT devices will increase. Artificial intelligence and neural networks are becoming pervasive, and a distributed approach all the way to the end nodes of the network will be required.
To fulfill all of the requirements I have outlined, we need a new, optimized approach to engineering. The hierarchical approach has many inefficiencies that must be eliminated, and this will happen only by blurring the lines of knowledge between engineering teams and disciplines. Each component to be designed in the context of system programmability and specific use cases requires multi-disciplinary engineering talent.
How can we this translate this new approach in our educational systems? What types of electrical engineers do we need for the future? Specialization will still be important, but once more, electrical engineering jobs will also require Renaissance scientists who are knowledgeable and agile across many disciplines. To begin to achieve this vision, here are some of the classes that should be available to the next generation of engineers:
- A strong foundation in scientific fundamentals including basic and advanced mathematics, physics, and chemistry. Mastery of statistics is also very important.
- A strong foundation in electrical engineering, including fundamental analog and digital design, all the way to transistor-level operation, manufacturing, and testing.
- Basic knowledge of coding (software development) independent of specific technology interests. Engineering students need to learn a scripting language (Python expertise is currently the strongest need) and a programming language (C and/or C++).
- Classes that teach engineers how to build things, primarily based on microcontrollers or even mechanics.
- Coursework in wireless communications, including RF design, is a plus. There is not enough material (copper, aluminum, silver, and plastic) to wire the entire world, especially for “things.” In the future, almost everything will be wireless, and we will have to communicate effectively in a crowded spectrum.
It is not only important to understand how to build things but also how long it takes to build. The best approach is to support as many internships as possible, ideally covering many aspects of engineering.
In essence, we must accelerate engineering classes that leverage the ecosystem of development tools and application building blocks that are now available to build connected things. At the same time, we must continue to build a strong foundation of basic science and engineering.
By Ramanuja Konreddy, sr. product marketing engineer, Microchip’s 32-bit Microcontroller BU
Most useful included Computer architecture, VLSI Design, Device physics, and Microcontrollers. 
No comments:
Post a Comment