Overcoming six challenges of UX design for IoT

July 26, 2017 Arup Barat

Growth of the Internet of Things (IoT) continues to skyrocket, with analysts predicting that by 2020 annual revenues could exceed $470 billion for companies selling hardware, software, and comprehensive solutions, with the market forecast to grow from an installed base of 15 billion devices in 2015 to nearly 31 billion devices by 2020. While connecting smart devices is catching on both at home and at work, there is still room for growth and improvement. One such area is user experience (UX). Creating a strong user experience is key to turning the IoT from the current hot trend into the next phase of computing.

UX design is important because it determines how a user interacts with a system. While UX design practices for web and mobile have become standardized, designing for the IoT is a new realm that involves a wide array of different devices, thus making the design process more complex. Here we’ll look at six challenges UX designers face when designing for the IoT and solutions to those challenges.

1. Heterogeneous systems

IoT solutions need to be able to handle many types of data from an array of devices and display the data on a variety of user interfaces (UIs). For example, an IoT solution at a solar power plant might collect data from three or more devices or sensors, like an inverter, weather station, and grid pricing. The inverter and weather station would likely use a processor and wireless communications chip to collect and transmit data through a gateway to the cloud, while the utility company would use an application programming interface (API) to make grid pricing data available. Once all these different types of data are aggregated, the end user needs to be able to easily read it on any interface, like a phone, laptop, or on-site screen. Additionally, each type of user – owner, financier, or technician – needs access to different types of information. Designing for so many types of data, devices, interfaces, and end users is what makes UX so challenging.

The heterogeneous nature of IoT systems is intensified by the fact that each system is constantly changing as devices are added or replaced in the field. For example, if the solar plant mentioned above expands to add a second inverter from a different vendor, the IoT system not only needs to work with the new data collection infrastructure, but seamlessly mesh with the end user interfaces as well. Designers need to ensure that their UX is flexible enough to adapt in situations like this.

Another challenge created by this diversity of devices, data, users, and more is creating a unified feel across all interfaces to enhance each user’s experience with the solution. To make this happen, designers often have to work in a more basic program like Linux to code for each interface properly. For example, a user viewing data on a smartphone expects a beautiful display; while that same design likely won’t work on a simple screen on a piece of machinery, the general concepts and experience should translate.

2. Focus on hardware

When it comes to choosing IoT hardware, buyers tend to focus on cost, technical specs, and software compatibility while ignoring user experience. But hardware can have a big impact on UX. It’s crucial that the appropriate sensors, processors, and communications modules for edge devices are chosen properly, as they ultimately determine how a user is able to interact with a device. For example, if a company selects a processor because it’s the cheapest option, it may have a slow response time or run out of battery quickly. This will significantly diminish UX in most cases.

Enough MIPS/RAM/ROM

  • Running a real-time operating system (RTOS) or time-based scheduler would be sufficient for most applications, but limits the ability to expand

A variety of serial Interfaces

  • Serial I/O is needed to communicate with peripherals like GPS, cellular modems, etc.

  • For expandability, more than the minimum serial interfaces are required

A/D and GPIO

  • A/D and GPIO is not a must have, but these are features that may be needed in the future

Durability

  • Systems targeted at hobbyists and minicomputer applications may not be rugged enough for field use

  • None of the systems, as built, may be fit for extended use in the field

Manufacturability

  • Meeting field use environmental conditions (heat, vibration, dust, moisture) may require a custom board, regardless of what is chosen

  • Some systems use components that are difficult to obtain, especially for smaller companies/volumes

Typical hardware selection criteria

Designers can create the most beautiful interface, but if end users are trying to use it with incompatible hardware, the user experience will be a poor one. When selecting hardware components that will make up an IoT solution, buyers need to keep the end UX in mind from the beginning.

3. Getting connectivity right

Choosing the right type of network for an IoT system is also very important to UX. There are many networks that support the IoT, all offering different things – faster speed, lower cost, data limits, and more (Figure 1). Each IoT use case requires a different connectivity solution, and selecting the right one plays an incredible role in UX. In some situations, such as autonomous vehicles or volcanic activity monitoring, users need super low latency and minimal data loss. In other situations, such as solar generation or tide monitoring, those issues are less important. It doesn’t matter how seamlessly the UI works; if the data doesn’t reach its end point when it’s needed, then the system has failed.


Figure 1: Common communication technologies with features and capabilities (source)

When choosing a network, it’s crucial to remember that many industrial IoT solutions are remotely located and use basic devices that easily lose connectivity or miss data points. While some network options are more reliable in these situations, UX designers generally need to implement a way for field devices to respond after going offline. Options include a solution that smooths over missing data, clearly notifies the end user that the device has gone offline and will be missing data, or showing that the data is in process until the device is back online and cached data can be sent. The designer’s solution should be based on the end user’s expectations as well as the situation and use case (Figure 2).


Figure 2: The Nest UI clearly shows the user if a specific device is offline.

4. Design for all modules

IoT solutions are typically built on an IoT platform that provides a foundation for many of the necessary components – hardware, data collection, application development, actions, rules, analytics, connectivity, and more – to work together seamlessly. From this base, designers can build out the full solution. However, each component plays a vital role in the end UX, so designers need to understand how each piece works and what impact it has on the end solution. Because many of the components aren’t visible to the end user, UX designers can easily focus only on the parts everyone can see (Figure 3). This would be a mistake.


Figure 3: Visibility of IoT UX elements

For example, end users don’t see the rules engine. In a smart manufacturing facility, a machine operator may rely on a real-time notification to alert him that the machine needs to be stopped for maintenance. That notification is sent as the result of a rule being triggered by the rules engine. The operator likely doesn’t care how the rules are built, but if the rules aren’t set correctly or notifications aren’t delivered in a timely fashion, the operator will suffer. If a UX designer doesn’t pay enough attention to the rules while designing, notifications may not be sent when expected or displayed in an awkward format, thus diminishing the UX.

5. Working with many vendors

Third-party vendors supply many of the components needed to build out an IoT solution, and the sensors, processors, controllers, platforms, and applications used are likely not coming from just one supplier. It can be difficult to make those pieces work together and create a seamless UX. Additionally, companies will likely continue to replace old hardware or update software and operating systems (OSs). All of this can lead to incompatibilities, which can make a positive UX more difficult to achieve.

Overcoming interoperability challenges can manifest in several ways. For example, if the user is forced to switch between two different programs or screens to access data from two different devices or control those devices, the UX will be inherently poor. This is commonly the case in the smart home where a user may operate the smart thermostat with one app and the smart security system with another. At the extreme, a user may have two brands of video camera and motion sensors as part of a security system that can only be accessed using two apps. This clearly isn’t ideal, and it’s likely to cause the user to abandon the IoT solution altogether. Because there are so many manufacturers in the consumer space and many ecosystems are closed so data can’t be brought to a central location, it can be more difficult to achieve a seamless UX in the smart home than in the industrial space.

Figure 4: The abundance of smart home apps that don’t work together leads to a bad user experience (source)

6. Building trust between user and machine

For an IoT system to be used to its limits, users need to fully trust the underlying data. For users working in a highly sensitive environment, like a nuclear plant or monitoring a volcano for city safety, trusting a dashboard on a smartphone with numbers popping up or a good/bad dial might be difficult. Right or wrong data can make all the difference in a user’s decision in a high-pressure situation, which is why UX designers need to build trust into the experience.

Designers can do this in several ways. The solution can provide status updates, allowing a user to understand exactly what’s happening with the system and feel more comfortable when making any decisions. The IoT system dashboard could also allow users to take a deep dive into the data with drill-downs that help the user understand why certain things are happening and feel comfortable. The software doesn’t need to force these details on users, but can allow users to engage and get more comfortable with the data that is being provided. By providing this type of transparency, a good design can improve UX, particularly in stressful situations.

When creating an IoT solution, it’s necessary to focus on the big picture of what the system will do and how it will get there, but it’s equally important to consider UX design and incorporate it into every aspect of the project. To really get the most out of an IoT system, a seamless UX is vital. The only way to overcome the challenges in this new space is by focusing on UX design from the outset.

Arup Barat is the co-founder and Chief Commercial Officer of infiswift, an enterprise Internet of Things (IoT) platform company based in San Ramon, CA, where he guides commercialization activates at infiswift through product, marketing, sales, BD and customer success.

 

About the Author

Arup Barat is the co-founder and Chief Commercial Officer of infiswift, an enterprise Internet of Things (IoT) platform company based in San Ramon, CA, where he guides commercialization activates at infiswift through product, marketing, sales, business development, and customer success.

More Content by Arup Barat
Previous Article
Automotive ECUs: Architecture considerations to implement secure software updates over the air
Automotive ECUs: Architecture considerations to implement secure software updates over the air

With increasing recognition that our cars are evolving into rolling data centers, manufacturers are faced w...

Next Article
Video compression: Lost image resolution
Video compression: Lost image resolution

If your camera and recorder, storage, or displays are more than 500 feet apart, compression will likely be ...