Perhaps you’ve heard the story. A century ago, supposedly no less than the United States Commissioner of Patents had remarked, “Everything that can be invented has been invented.” How ironic! In fact, what Commissioner Charles Duell actually said back in 1902 was quite the opposite. He stated, “In my opinion, all previous advances in the various lines of invention will appear totally insignificant when compared with those which the present century will witness.” Most of us would probably agree. What “wonders” will those be? What new factors may influence the nature of how we approach problems and develop solutions as engineers?
Of course, attempts to predict details of the future with any degree of precision are fraught with folly. But history does provide us with a few insights into what might remain the same. For example, unforeseen and previously impractical technologies are certain to foster new industries; and key market opportunities will correspondingly accelerate the discovery, development, cost-reduction, and ultimate widespread adoption of those technologies. This is as true today as it was when the first Roman aqueducts were built, or when Watt and Boulton improved upon Newcomen’s steam engine and helped catalyze the industrial revolution. A more recent example is autonomous automobiles—they seemed to suddenly leap from the realm of sci-fi into reality in less than a decade.
We see every successive generation of engineers enter a world that innovates faster, by collectively leveraging tools and knowledge from their predecessors. This too, seems unlikely to change. But the tools in our hands today, like the proliferation of low-cost 32-bit microcontrollers, pervasive wireless connectivity, novel sensors, and Internet of Things (IoT) functionality present a possible inflection point for the practice of engineering. Market opportunities and needs aren’t confined to new categories of applications either—we’re re-imagining the potential utility of “common” products as well. When I was a young boy, I delighted at the chance to take apart old wall-mounted thermostats—a prized glass vial of mercury could be found inside, providing hours of reckless entertainment. Now they are being replaced with smart thermostats like Nest. Mercury switches are on their way out. Look inside a Nest thermostat and you’ll find a 32-MHz ARM Cortex-M3 MCU running Linux, a 24-bit color LCD screen, along with sensors for temperature, humidity, far-field activity, near-field activity, and ambient light, 802.11b/g/n and 802.15.4 WiFi, and a lithium battery. Even the humdrum doorbell is being transformed into an app-linked video security device with products like Ring and others.
As these trends continue, the products we’ll be developing may need to address ever-present security/vulnerability concerns. They may offer capabilities based on cloud computing and predictive analytics, be integrated with smartphone apps and services, collect and interpret data from a myriad of sensors, be mindful of their power consumption, control and communicate with other vendor’s devices, and provide a seamless end-user experience.
Factors That Can Change the Future of Engineering
The first is the need for sensor fusion and data abstraction, “at the edge.” Why? Sensors are everywhere—after all, they are the eyes and ears of smart and not-so-smart products. At a high level, the constant stream of measurements provides a flood of data, and often a “deficit” of useful methods to transform that data into meaningful, actionable, and potentially profitable information. In essence, IoT proliferation is presenting us with more data than we know what to do with, and consuming excess bandwidth resources (and energy) in the process. This suggests we’ll need to understand and creatively leverage distributed artificial intelligence (AI) to perform data abstraction at the edge in robotics, medicine, factory and home automation, agriculture, and for other sensor/mechatronics-based applications. It won’t be enough to design-in the latest set of sensors. We’ll need to think in terms of information, not data. AI won’t be relegated to academia, consumer devices, and personal assistants; nor will it require expensive dedicated processing hardware or cloud resources. AI will follow the path of low-cost 32-bit MCUs. It’s destined to be an essential component in many products and a frequently used tool in our design arsenal.
Just-in-Time Approach for Gaining Knowledge
Another factor may be the requirement for us to balance a reliance on our current “traditional engineering” skillsets with a just-in-time approach for gaining knowledge in other domains—biology, botany, chemistry, optics, music, art, etc. Think of this as an evolution of the just-in-time/best practices used in manufacturing; it’ll play a key upstream role in a product’s market definition, concept development and design phase.
Let’s say you’re involved in designing LED architectural/industrial lighting systems and are very experienced in this domain. You understand your end-customer’s correlated color temperature (CCT) and intensity requirements. You use familiar tools for thermal analysis and optical modeling. Now you or your company is interested in entering the horticultural lighting industry. An LED module is an LED module, right? Perhaps, but the “end customers” in this case are plants. From an electrical perspective a properly designed horticultural lighting unit isn’t radically different from other lighting products. However, the LEDs may need to be selected to provide the optimum wavelengths (colors) and output power, which generally vary by plant type and its developmental growth phases. In fact, even the unit of measurement for optical power will be different. Instead of the familiar foot-candles or lux (illuminance—the quantity of light falling on a given area), you’d probably be presented with a target value for photosynthetic photon flux density (PPFD), which is the number of photosynthetically active photons that impinge on the surface of the plants each second. Your electrical and mechanical design experience will not be enough. You’ll need on-demand practical knowledge in some aspects of plant science, ideally for the particular target plant.
A further example is the task of upgrading an existing MCU-based product like a home appliance, to enable IoT functionality. The hardware changes may be relatively easy because so many off-the-shelf communications modules are available for LTE, WiFi, Bluetooth, LoRa, and so on. Indeed, hardware will not be the design challenge. It might be security: determining the business-risk and cost of a data breach, uncovering and addressing all of the potential attack surfaces from your device up through to the cloud, preventing device “spoofing” with proper authentication, and how to deliver secure over-the-air firmware updates. Again, you’ll need on-demand practical knowledge.
Incorporation of Art in EngineeringAs we will increasingly live and work in a world where unrelated devices and processes must interact with each other, the future of engineering will necessarily encompass mechanisms to acquire just-in-time knowledge across a wide range of domains and disciplines. Engineering students will be taught the value of finding and validating this information quickly. Agile knowledge acquisition could be a major determinant of their future success.
The third factor will be the increased significance of art and graphical design in our practice. This correlates with the trend to enrich students via STEAM (Science, Technology, Engineering, Art, and Math) teaching concepts. As new engineers, our sons and daughters are viewing these as entirely synergistic disciplines.
Where does science end, and art begin? For some anthropologists (and primatologists), one common characteristic of a tool is that in the hand (or claw, beak, what have you) of the user, they function as “transparent extensions.” When Jane Goodall first observed chimpanzees gathering sticks, stripping off the leaves and using them as “fishing rods” in insect mounds, she realized the sticks were transparent—in other words, when used for insect fishing, the animals didn’t view them as sticks any longer, they were functional extensions of their hands and fingers.
Now let’s go back to the question of science-art boundaries. There’s a not-so-subtle shift in product design, especially in the consumer space, for immersive interaction experiences that reflect the user’s lifestyle, or the activity they are focused on. The device itself might be secondary. Samsung’s new “The Frame” series of TVs is a recent example. When the TV is not on, it becomes a framed work of art. Users can even browse and download new art from Samsung’s “Art Store.” Those TVs have a degree of transparency; just as the stick isn’t viewed as a stick, the TV is no longer a TV when it’s off.
Design aesthetics and functional, thoughtful design really do matter. It’s evidenced by the most common interface people use for their smart appliances—the app on their phones. The general (and long overdue) trend is for smart devices to conform to the user’s expectations, not the other way around. This provides a measure of transparency for the hardware and firmware. Future engineers are destined to embrace human-centric design principles. Emphasis on contextual awareness and understanding the user’s implied intent may be placed on par with circuit performance; it can be an important differentiator in the market and help establish a compelling brand preference.
Conclusion
What will the future of engineering look like? It will place a much higher value on information rather than on data, and leverage AI at the edge when necessary. The ability to quickly find, validate, and leverage interdisciplinary knowledge will be a critical challenge for many engineers and engineer-entrepreneurs. And finally, as we encompass art and human-centric design principles both in engineering education and in our everyday practice, our ability to positively influence our lives and our world will continue to reach new heights.