A 3D SOLIDWORKS CAD model of a Haiku fan. A Haiku fan embeds the fan light and uses airfoils instead of planks to move air around the room. The result is a sleeker, more optimized design. A sleek design optimized to perform its task efficiently should be enough reason to buy the product. IoT is just a cherry on top. Engineers need to understand that this is often the case. Start with a good product and add IoT. IoT doesn’t make a product good; it just adds to it. (Image courtesy of Haiku Home.)

Often, the best-designed products just blend into reality and are taken for granted. But that doesn’t mean there isn’t room for improvement. When it comes to fans made by Haiku Home from Big Ass Solutions (yes, that is its real name), the company uses its expertise of blowing air around to not only optimize the design of fans, but also to add them to the Internet of Things (IoT).

Starting with a Good Disconnected Product Is the Key to a Great IoT Product

“If you don’t get the product right then, you might as well not bother with the smart piece,” said Landon Borders, director of Connected Devices at Big Ass Solutions. “If done correctly, adding connectivity to the product is just an incremental improvement.”

He explains that the start of any good IoT product is the product itself. You can’t just slap connectivity to any old design and call it a day. You need to think how you have improved the product for your customers. Not all products lend themselves to the IoT. Similarly, no amount of connectivity features will improve a bad design.

“A lot of the challenges with designing an IoT device comes down to not having a playbook,” said Borders. “A lot of this is very new and we’re all learning as we go. But it starts with the product. I see so many smart devices out there, and they are really not solving any problems for anyone. I shake my head sometimes. So one of the challenges is learning what customers want and translating those needs into specifications and requirements.”

When it came to classic fan designs, there were many things Haiku Home discovered it could improve upon from a user perspective. First, take those noisy motors. Borders explained that Haiku works on brushless motors that reduce the noise and improve the life of the product.

Next, the dangling lights and chains often associated with ceiling fans had to go. The chains were always confusing to use, they fail or break frequently, and they are difficult for those who aren’t especially tall to use. And the dangling lights could become a literal headache inducer for those who are tall. These interesting design changes were all crafted in a SOLIDWORKS CAD model.

Control for the fan went to an IoT app, while the lights were replaced with LEDs that are built into the fan’s sleek frame.

To really see how much energy Haiku Home puts into the design of its IoT products, before even adding IoT, look no further than the blades.

Traditionally, the blades on fans are a little more than a plank. However, Haiku Home designs its blades into airfoils. The company even performs computational fluid dynamics (CFD) simulations to optimize the fans’ design. These simulations are also used by Haiku Home’s sales team to ensure that a fan is properly sized for a room and the contents within it.

“We also simulate the air movement and how light is distributed into a facility,” said Borders. “Our design engineers can simulate the air movement and then calculate how a specific pitch or angle on a foil might produce a particular cubic foot per minute (CFM) [flow], and this is normally spot on when we get to the physical prototype.”

In other words, Haiku Home wanted to make a product that people would want because it was already an optimized stand-alone product. Adding IoT just became the cherry on top.

I Have a Good Product Design. How Do I Make It a Great IoT Design?

“A lot of people get caught up in the connectivity and interoperability piece,” mentioned Borders. “But If I were the one designing the product, I would think first about use cases and what the customer would want accomplished that they don’t have today. Ask what sucks about a product, and figure out how to make it suck less. But in our case, we blow air.”

In fact, Borders noted that adding the connectivity to a product itself isn’t the big challenge when designing IoT products. After all, many products already incorporate electronics.

“Electrical engineers, mechanical engineers and industrial designers have been fighting over space in products for years, and that hasn’t changed,” explained Borders. “The only exception to that might be antenna placement, which often leads to material decisions."

Therefore, answering why your product needs IoT is a lot harder and more important than answering how you will add the connectivity in the first place. As an example, ask yourself, will your product need to be a part of a larger IoT ecosystem? Will the user experience of your fan improve if it can talk to the stove and know to speed up to help cool the kitchen while you are cooking?

“It’s one thing to be connected, but it’s another thing to have something to say,” said Borders. “I think about these things in terms of ecosystems of products. Your blender and your fan and thermostat really don’t have much to say. They can be connected, but they don’t need to be connected to one another. When I think of our ecosystem of comfort and energy conservation, I’m thinking about products that have something to say to one another, and when they do, they can sense environmental conditions and react accordingly.”

Communication between products on the IoT can become complicated. Take ecobee’s smart thermostat. It has coin cell battery sensors located around the home and a wall-mounted thermostat with a battery backup. Due to these power limitations, the thermostats will need to be strategic when communicating data. In contrast, Haiku’s fans are plugged in and do not have a limited power supply. As a result, they are continuously connected.

“For us, that means we can only pull the smart thermostat for information at certain intervals and because of that there is some latency. This latency isn’t that perceptible to our customers, but you do have to think [during the design] how this latency affects the user experience.”

I Have a Great IoT Design. What Do I Do with All This Big Data?

“People see value in data, but they don’t know what it is,” remarked Borders. “Oftentimes, they will jump to ‘How do I monetize this?’ But that’s not how we look at it. For us, the data gives us insights into how our products are being used at a macro level.”

Borders joked that they don’t have an employee looking at the data come in like some operator in the Matrix movies. Instead, Haiku Home aggregates the data, sterilizes it from personal information, and looks for patterns in customer usage.

“Looking at the data at an aggregate view gives you an idea of how products are used and how you can make them better and create new features for customers,” explained Borders.

For instance, Borders explained that by looking at the data, the company discovered that over 90 percent of the time users were using their fans at 60 percent of maximum speed. In other words, the vast majority of the fans were over-engineered for their purpose.

As a result, the company was able to design a fan with a smaller motor and lower profile without affecting the performance expectations of users. The added benefit is that this design translated into cost benefits for both Haiku Home and its customers.

I’m Collecting This Useful Big Data. Now How Do I Secure My IoT Connection?

Security is dependent on the product and the service it provides. As a result, engineers need to take into consideration the product and what might be on the network it is connected to. They then must theorize what might go wrong in the event of a hack.

In other words, security means something different for those who are designing IoT deadbolt locks than an IoT fan. But once that fan connects to the deadbolt, then things can get fishy.

“People look at the scenario at Target, which was a very high-profile security breach,” said Borders. “There was a lot to lose there, and I can’t believe how there was a loophole where the hackers hijacked the air conditioning and somehow ended up with a bunch of credit cards.”

“That was a massive failure on many fronts, not just smart devices,” he added. “Security means a lot of different things, and you need to think of the consequences if someone hijacks the control of a smart [device]. They might turn on and off a light and it might be annoying, but if someone hacks a security system, the repercussions are a lot greater. So we look at security at a bunch of different layers at the security stack.”

Another thing Borders notes is that security and privacy are not the same thing, although they are certainly connected. When collecting customer data from the IoT, it is important that the data that goes to sales is different from the data that gets to the design engineers. Engineers don’t care about your name and zip code. They care about how you use the product. As a result, it is best to use firewalls to separate this data to ensure the privacy of the users and to ensure that their personal information is secure.

To help ensure that no alarming errors exist in an IoT product’s security, Borders suggests that a company use third-party security audits on a regular basis. The last thing you want is to release a product that accidentally saves the Wi-Fi log-in information in plain text somewhere in the system. Mistakes like this can happen if you are not careful and you don’t look into your system. Performing regular audits of the product and company will help reduce the likelihood that these vulnerabilities will occur. So get in contact with a good white hat.

Finally, Borders suggests that security doesn’t rest solely on the engineers creating the products; users need to think about their own security as well. He suggests that users should use Wi-Fi Protected Access (WPA) 2 encryption for their networks and ensure that their passwords are complicated enough to limit hacking. He explained that it’s difficult for hackers to crack such an encryption. And if they can’t get into the network, they can’t control anything. He said, “WPA 2 is very important. That’s a layer of security that you just can’t overlook. So there is some fear mongering in the media.”

About the Author

Shawn Wasserman (@ShawnWasserman) is the Internet of Things (IoT) and Simulation Editor at ENGINEERING.com. He is passionate about ensuring engineers make the right decisions when using computer-aided engineering (CAE) software and IoT development tools. Shawn has a Masters in Bio-Engineering from the University of Guelph and a BASc in Chemical Engineering from the University of Waterloo.

CMU's connected Coke machine

Sometime in 1982, a seemingly innocuous conversation occurred in the Carnegie Mellon University (CMU) Computer Science Department. Programmers logging long hours in the labs were frustrated by their long walks to the Coke machine, only to find it empty or, much worse, filled with warm soda.

You see, around the halls of CMU, caffeine wasn’t just a substance, it was a driving force. Something needed to be done about these errant trips to the machine.

Soon, conversation spun into action. The CMU Coke machine was filled with a number of microswitches and connected to the Internet. And soon, its contents and their temperatures could be found by anyone who could reach the IP address: 128.2.209.43.

Although it didn’t seem like much at the time, the moment CMU’s Coke machine came online, the Internet of Things (IoT) was born. Within a generation and a half, IoT would transform into the Industrial Internet of Things (IIoT) and become a driving force for business and manufacturing innovation across the globe.

What Is the IIoT?

IIoT (sometimes referred to as “Industry 4.0”) was first coined as a way to talk about the union of big data, machine-to-machine communications, machine learning and sensor information. The kernel of the IIoT idea was that the more sensors you could pack into a machine, the more valuable data you could get about its performance. Over time, tracking this data could give engineers insight into the performance of a machine and when it needed to be serviced, exchanged or reloaded with material, among other things.

Expanding this vision out a bit further, a manufacturing landscape filled with sensor-laden machines could link warehouses with manufacturing facilities and shippers in a seamless electronic conversation.

Today, the vision of IIoT has extended well beyond its initial idea and the notion that it would be a manufacturing-side enterprise. Most proponents of IIoT see a future where all types of machines and sensors, whether they’re industrial or consumer, are communicating with one another. This can offer predictive prescriptions and solutions to problems such as downtime, supply shortages and overages, consumer demand and much more.

What Are the Benefits of the IIoT?

Although increased operational efficiency is one of the principle benefits of a broad IIoT, the paradigm might also have other knock-on effects across industries that aren’t immediately apparent. One of the biggest impacts IIoT may have is that it will force companies to develop products in a different manner, focusing on software and experiences rather than product redesigns.

Today, consumers aren’t as interested in buying new models of products as they were in the past. Instead, they want a product that will grow with them and provide an evolving experience that can be updated and upgraded via software as time passes. With the IIoT, companies will be able to read more into consumer demands by reaching the data being generated by consumer sensors. This data can then be translated into product innovation. As companies become better adapted to this manner of product development, updates and upgrades may start to appear faster and faster. It might even be possible for companies to understand their customer’s wants so well that they can develop new user-specific product lines for more immersive and satisfying experiences.

Heading back the industrial space, IIoT could spark a change in the way humans and machines interact with one another, bringing them closer and shattering the notion that “the robots have arrived to vanquish the human workforce.”

According to the World Economic Forum’s analysis of the IIoT in 2015, human-machine workforce collaboration will be a key factor in improving workplace safety and diminishing workplace errors and more. According to the report:

“This new blend of labor combines human flexibility and contextual decision-making with robots’ precision and consistency to deliver better output. With its recent acquisition of Kiva Systems, for example, Amazon now operates one of the world’s largest fleets of industrial robots in its warehouses, where humans and robots work side by side, capable of fulfilling orders up to 70 percent faster than a nonautomated warehouse. While robots perform picking and delivery, human workers spend more time on overall process improvements such as directing lower volume products to be stored in a more remote area.”

In another example of successful human-machine interaction, the World Economic Forum’s report highlights how Marathon Oil employees working at its refineries wear multigas sensors that detect any exposure to harmful chemicals. With these sensors constantly moving around the facility, plant managers can get a high-resolution picture of the air quality and safety of the refinery. If levels of gas become dangerous, workers can be evacuated from an area quickly. Because of the machines they wear on their bodies, Marathon’s employees are much safer.

What Are the Hurdles for IIoT?

The first large hurdle before the IIoT is breaking down the language and protocol barriers that exist between machines. Due to the fact that manufacturing systems have always been developed in proprietary silos, it’s often the case that interoperability between differing machines is difficult. Creating a translator for machines, or a universal standard language for machine communication, will be critical for the IIoT.

Although a universal language for machines is a major hurdle to the success of the IIoT, that task is dwarfed by the need for incredibly tight security. If the Stuxnet controversy of yesteryear is a bellwether of what hackers can do with an industrial system, then I can image the likes of the oil and gas industry, utility companies, medical facilities and others will do much more than tap the breaks on IIoT if there’s even a hint of cyber insecurity.

With that in mind, its incumbent on those developing IIoT to build robust, multilevel security features, system checks and the like into the DNA of the project.

Finally, data propriety could also be an issue for those braving the early IIoT wilderness.

If the idea behind IIoT is to improve customer outcomes, production cycles and decision-making by leveraging massive amounts of data, it makes sense that companies may have to develop new methods for sharing data with other firms. This might mean that new encryption technologies, business models and hardware need to be developed to facilitate a broad exchange of data. It’s possible that third-party businesses might pop up to facilitate this type of trusted communication between firms, but if not, the prospect of developing or joining a robust IIoT infrastructure could appear exceptionally daunting to companies large and small.

Where’s the Industrial Internet Headed?

At the moment, the industrial Internet is still in a protean phase, and there are a number of concerns that surround the project. But in a few years, we might begin to see glimpses of what Larry Smarr, director of the California Institute for Telecommunications and Information Technology, described as a “sensor-aware planetary computer.”

Although that notion may still seem fairly blue sky, just take a second and look around you. Imagine how many devices in your home, office or even pocket are loaded with sensors capable of sending information back to complex machines and analytic engines. With that type of granular data available to industrial, civic and commercial interests, it’s only a matter of time before Smarr’s idea becomes good and the industrial Internet becomes a fluid machine delivering goods and services and possibly innovation on a perfectly scheduled basis.

What’s even more incredible is that all of this—the machine intelligence, the cross-industrial connections, the ability to control devices from half a planet away—all came about for the want of a cold Coke.

About the Author

Kyle Maxey is a mechanical designer and writer from Austin, TX. He earned a degree in Film at Bard College and has since studied Mechanical and Architectural drafting at Austin Community College. As a designer Kyle has had vast experience with CAD software and rapid prototyping. One day he dreams of becoming a toy designer.