By the end of 2017, it was projected there would be more than 8.4 billion connected devices in operation around the world. The ways a new product can connect and interact with other participants in the Internet of Things are growing exponentially. The challenge for connected device developers lies in how they decide to connect their device from a near endless and evolving set of technology choices. Some of the most essential decisions an OEM must make involve device communications technology, technology provider, platform service level and specific services to utilize, and the user interface device.
Decision 1: The Core of Connectivity
The beating heart of all connected devices is the underlying software that enables connections. Beyond connecting the device to the Internet of Things or smaller, contained networks, this software has the potential to harvest and present data about the device or its function to its users via a pipeline to the cloud. This pipeline has traditionally been via a hard-wired communications protocol, like Ethernet or USB.
More recent technologies can deliver this data via wireless chips that send data to the cloud via a set of wireless communication protocols, like Bluetooth, WiFi, or ZigBee. The choice of chip technology and communication protocol in itself involves myriad options, which provide varying degrees of flexibility and compatibility with cloud infrastructure providers. Some allow data to be retrieved from the device on-demand. Some allow data to be passively harvested. The needs of the full connected device system should be carefully considered prior to embarking on a full product development.
Decision 2: Connected Devices and the Cloud
Developing a connected device allows an OEM to leverage the device’s ability to generate massive amounts of data that can be analyzed to improve system performance and customer experience. The cloud can be viewed simplistically as the housing for that data, but it offers much more than that. For example, data can be filtered or sent through an analytics module prior to reaching the end user. Does the data need to reach the user in a time sensitive manner? If so, then it should utilize a cloud caching service for data transfer. Does the data need to be stored for a finite period? If so, then a cloud storage solution and associated services should be used.
Many cloud providers offer services to store the data in cloud-based servers using patterns of traditional SQL tables and relational databases—a wise storage choice when persistence of data is a top concern. The data could also be stored in BLOB or NoSQL format for faster retrieval.
Each design decision results in a different cost structure. For example, if the connected device needs both the cached method of fast data retrieval and the slower, traditional model, the data traffic through both channels will be charged on a “volume of data usage per channel” basis. The size of the data storage will also have an impact on cost; thus, the frequency at which data is stored and the duration for which the data is needed will drive up cost.
Off-the-shelf packages can reduce development time. But the trade-off here is that data usage through that channel ends up in a higher cost tier. In contrast, user level controls and authentication can be developed separately upon the base framework to keep data cost levels lower, but this separation usually carries with it a longer development period. Custom user level controls are typically more flexible and required in non-standard scenarios. These are but a few of the trade-offs in considering these controls that have cascading impacts on usability in later stages of development.
Decision 3: Which Service Platform Providers Is Best for Your Company?
All the cloud services mentioned above can be provided by literally thousands of service providers. These range from startups to large players in the cloud services industry. The largest providers offer the most flexibility and customizability; these are companies such as Amazon Web Services, Microsoft Azure, IBM Watson, or Google. Medium players such as Bosch, GE, Hitachi, Huawei, Intel, Oracle, or Salesforce often specialize in servicing particular industries.
Startups such as Exosite, Ayla, or ThingWorx are less flexible but tend to offer off-the-shelf packages that work directly with their module. This eliminates product development time at the expense of flexibility and feature offerings. The design choices within these companies’ offerings are pre-made by the service provider and limit the scope of how they can be used within a given connected device ecosystem.
Decision 4: Choosing the User Interface Device
Here, the number of options available to a product developer grow exponentially. The device could interact with an iPhone App, Android App, website, or a custom display or touch device that doesn’t rely on supporting off-the-shelf devices. Each one of these is an independent development effort.
Some development platforms enable parallel development of both iPhone and Android applications using the same source code, but they are in the infancy in their ability to produce a highly customizable application across both platforms. In general, if a client wants their product to view and manipulate data on all platforms, this results in a nearly equal sized development effort on each platform, multiplying the costs of product development.
In summary, being an “expert in connected device technologies” is like saying that one is an “expert in sports.” It’s near impossible to be an expert in all aspects of all professional and collegiate games. For the connected device developer, the challenge lies in defining what device technologies and in what aspects of those technologies one can claim expertise. It also requires knowing how and what to seek out when one requires knowledge or development within a new environment.
The user needs and connected device ecosystem need to be well defined before embarking on full product development. All elements of the ecosystem require consideration in concert with the product requirements, target cost, and potential market for the product. There are many combinations of tools and technologies that can achieve a desired result. Choose wisely, as the wrong combination will result in inflated costs, extended timelines, and decreased performance, usability, scalability, or functionality.
Need help with the design or development of your connected device? See how our designers and engineers can help.
Having trouble capturing and communicating the Service Value of your connected products to customers? Check out our new white paper about the 3 keys to capturing this value.