3/4 Free Articles leftRemaining

Does Hardware Even Matter Anymore?

June 09, 2015
Loading...
JUN15_09_185760494

We are in the midst of a technological revolution that is every bit as profound as the impact of cheap computing power, but it’s subtler and harder to notice. It will ease the way for companies launching and updating digital products, but it presents steep new learning curves that companies will have to master if they are to be successful.

What I’m referring to is the migration of functionality from hardware to software. In more and more businesses, physical objects are no longer the primary basis for innovation and differentiation. They come second to innovations in computer code.

Managers are well aware that Moore’s Law, the idea that the number of transistors on a practical-sized chip doubles every 18 months, has brought us a bounty of cheap computing power, leading to smartphones, tablets, fitness trackers, cloud-based services like Facebook and Uber, and on and on. But I’ve found that they’re less cognizant of how software has transformed other fields that we traditionally think of as hardware-based.

Consider, for example, how we convert and control electrical power. Think of the cubes we plug our iPhones into, the sensors that control our heating and lighting, and the motors used in tiny disk drives and the giant traction motors in locomotives. Modern solid-state power electronics got started in the 1950s, but rapid recent progress in power semiconductors, new power conversion topologies, and methods for controlling electric motors has brought us a plethora of small, high-efficiency, low-cost, and long-lived electronics subsystems for motion control. For a few dollars, designers can easily connect a computer to remember the seat position in your car. They can also replace the hydraulic power steering with a more-efficient electric power-steering system, or for that matter control everything needed to make that car autonomous — all it takes is software.

Insight Center

The biggest benefit from this trend is that you can incorporate more sophisticated control regimes into products. Old-fashioned analog controls require tuning and are expensive to manufacture. Software control allows you to plug in control schemes that would be almost impossible to implement otherwise. I recently rented a Volkswagen Beetle, and I noticed that when I opened the door the window rolled down just a little bit, anticipating the air pressure buildup that would occur when I shut the door. I got a nice satisfying door slam, and afterward the window rolled up. That would be really hard to do with analog controls, but with software? Easy. That’s one reason high-end cars have as many as 100 microcontrollers and 100 million lines of software running them — to power what Toyota calls “hospitality features.”

This style of “more electric” control by software also means big gains in energy efficiency. With electric power steering you draw power from the engine only when you need it, not constantly as with a belt-driven hydraulic pump. One automotive engineer told me that mandated fuel-economy standards were forcing manufacturers to replace engine-driven mechanical and hydraulic loads with electric. Lost in the furor over the lithium-ion battery problems with Boeing’s 787 Dreamliner were the efficiency gains from the plane’s “more electric” architecture. Boeing substituted electrically operated subsystems for traditional hydraulic and pneumatic power in key subsystems like flight controls, the environmental control system, landing-gear retraction, and braking. Not only do these subsystems draw engine power only when they need it, but more electric means less weight from hydraulic lines and ducts. Of course it also means a lot of lines of software.

The rapid uptake of smartphones has enabled manufacturers to rapidly scale up the production of sensors — GPS sensors, accelerometers, image sensors, capacitive touch sensors, all kinds of devices that help us measure the analog world — and connect them to our electronic world, where they can control things on the basis of what we see, hear, or feel. The sophistication and efficiency of these sensors are advancing rapidly, as you might expect given that their makers are supplying them for the manufacture of 60 million iPhones and even more Android phones each quarter. So it has become really inexpensive to add sensing to all kinds of devices — rear, side, and front vision on a car; an accelerometer to monitor your clothes dryer. The most innovative applications are probably still to come.

Harnessing all this technology — the computing, the motion control, the sensing — poses a huge challenge, but rising levels of abstraction are giving product designers the tools to meet it. By “abstraction” I mean the isolation of something’s essential properties so that it can be generalized and reused for wider application. Many software developers will tell you that the whole history of the software industry can be described by increasing levels of abstraction.

Abstraction allows product designers to conceptualize ideas at a higher level, which enables better and more innovative designs. It’s like using building blocks, adding the custom pieces, and then rapidly deploying them. If you need a standard building block that gives you internet connectivity, a camera, and a programmable computer, you can always use an iPad or a smartphone as a starting point. The advent of the iPad raised the level of abstraction for a whole group of hardware builders who formerly buried PCs in their systems. Notice all those new point-of-sale systems, or the remote-control apps that use them? They’re based on the iPad. Cloud computing abstracted away the whole provisioning of computing services for firms like Uber and Airbnb, as well as Nest and other hardware builders.

There are important implications for companies. For corporate leaders, one of the key lessons is that higher levels of abstraction shrink the entry barriers to numerous businesses — it seems that everyone can develop a new digital product. Companies need to be constantly on the alert for the next software-based product that might pose a competitive threat.

For product designers, the first implication of the software-replaces-hardware trend is that a much higher proportion of the value of a product will be in the electronics. The Boston Consulting Group estimates that the cost of the electronic parts will rise from 20% of the value in a typical automobile in 2004 to 40% this year. That means a major shift in the supplier network, with consequences that many are not prepared for.

It also means that much more of a product’s differentiation will be expressed in software. Over-the-air updates give firms the opportunity to add features, fix mistakes, or optimize performance, after the hardware part has been shipped, as long as the hardware design is robust enough to handle more demands than initially planned. When NASA sent the Curiosity rover to Mars, it discovered a software bug after the spacecraft had been launched. “Software can be updated, but hardware is fixed,” one of the engineers explained about the ultimate over-the-air update.

Software development will be more complex. As engineers take on more-complex control regimes, real-time software development and simulation tools will play a critical role in system designs. Complex-systems designers already know how to do this, but as usage becomes more pervasive, more firms will need to learn how to manage simulation and testing tools, as well as how to manage software complexity.

Finally, connectivity will assume a bigger role in the functioning and differentiation of products. Thus designers will have to take security seriously. For some applications like automobiles, manufacturers are putting firewalls between the infotainment side and the vehicle control and powertrain side. But for many new hardware devices that “live on the net,” we are entering a brave new world where security strategy is going to have to be a core design principle.

The software revolution will be a powerful complement to the cheap-computing revolution, and the opportunities for unique and innovative products are boundless — it’s just a matter of programming.


Willy C. Shih is the Robert and Jane Cizik Professor of Management Practice in Business Administration at Harvard Business School.


Loading...
Loading...