Let's wrap up this last section of the class on design of IoT devices. We talked first about designing for data, trying to look at the flow of data, the types and sources of data, the differences between big and small data, other ways to differentiate the data. Thinking about data in context, how it affects actions that we're trying to cause or effect, feedback loops where we work with users to provide data to help to make decisions and modify behavior, and progressive disclosure, where we focus on the key data that we want to present and then work our way into supporting data as it's needed. Some good references in this section, if you need to look into this further. We then talked a little bit about some broad design perspectives. Don Norman's emotional design approach for devices is an interesting one. We talked about the impact of interaction and placement of devices on their design. The concept of a visual brand language to give design its own personality, and to tie it to other devices in a product line. We talked a bit about the faces of product design, considering a product's story, the art and impact of the device, and then the science side of the device, what it has to support, how it's going to be created, etc. We talked about the importance of design consistency, which again, falls back into things like visual brand languages and some of the phases discussion. But it's important to realize that in any design, whether it's from step to step of using the device or whether it's across a family of devices, consistency is probably one of the most key elements to try and keep in front of us. Again, some good references here. We dug into something a little more crunchy around the use of Data Analysis in AI for IoT networks. Given that the devices have improved in their processing power, there's a tendency now to start looking at edge analytics to try and do more of the work at the device itself, rather than feeding it to a bridge or feeding it up to the cloud. To look at that, we talked about how we consider data that's in motion through that IoT network where that it might be at rest in a database, we talked about the value of analytics, how they're used. Then the uses of AI and machine learning, both the supervised and unsupervised flavors, and how those apply to IoT devices. Then the difference between the big data analytics that we would do with data collected and brought to the cloud versus edge analytics that we might do at the actual device. Again, some good references here. Design for wearables. This is really more focused on a specific subset of embedded devices, the ones that we keep on our person. The definition of wearables, the types of interactions that we have, we took a little detour into Bruce's Folly around his quest for keeping devices with him all the time. We talked about really what enables wearables these days, and the issues, and types of wearables that you might see, and then a little bit around how you would design for a wearable, and where these things are going, what the trends are, what the growth is, and what types of devices are most commonly worked on as embedded development projects. We then got into the design for voice user interfaces. These are all around us now with a Alexa, and Siri, and all the other voice-controlled devices that really weren't there 10 years ago. This is a relatively new area. We talked about the history, the definitions, the advantages and disadvantages of using these interface approaches. Especially considering that 10-15 percent of the population may not be able to use the devices in this manner. We talked about how you would design commands for these devices, the usability of a user interface that's voice-based, looked at some of the design principles, the issue of providing an avatar for a voice user interface, and the impact of the uncanny valley. Trying to make something that's just lifelike enough to really disturb someone that has to deal with it. The tools, the testing, and other issues that you might run into with user interfaces that use voice, we've reviewed all of that. Hopefully, the references here would help if you dig into this further for your own designs. To wrap up, the last assignment will be a peer review of the work that's been done for projects 1 and 2, the Qt and HTML projects. You'll get to peer review other people's projects. You'll have your own projects reviewed. That should be an interesting interchange and hopefully give you some perspective on how these designs come together. There is a quiz for this module. This is the last of the modules for the rapid prototyping section. Please feel free to reach out if you need any support around this last part of the class, and best of luck with it.
