The rapid adoption of intelligent IoT gateways has gone hand-in-hand with the proliferation of connected devices. Traditional data centers would struggle to cope with the tide of sensor-generated data and provide the interoperability required. Located at the edge, IoT gateways connect IoT devices to the cloud and bring to life any “smart” environment by providing connectivity and security while bridging different protocol adaptors with all the different types of cloud platforms — Microsoft Azure, AWS IoT, IBM Bluemix, etc.
These invaluable connection points work by “ingesting” data from IoT sensors. They transfer data to the cloud while at the same time receiving data from the cloud which they direct back to the devices. Intelligent gateways handle the different protocols to enable a seamless user experience, managing high-volume data flows and robust security.
Traditional gateways provided basic functionality. Today’s smart IoT gateways differ in a number of important ways. By performing the processing and analytics of data at the edge, they reduce latency and enable a whole host of critical use cases. They “edit” the data flow to pass on only what’s most relevant, and prevent the whole infrastructure from being deluged by data.
There have been IoT gateways for a number of years and they are valuable additions to the ecosystem. Nonetheless, how to deploy them — and when not to — is not yet widely understood. There are a number of factors to take into consideration when assessing whether it is right for an IoT initiative.
Designed for scale, performance and diversity
Intelligent gateways have become the go-to technology for IoT environments with many co-located edge sensors and where real-time responses to data are required. It is worth remembering that IoT is still in its infancy. As use cases and technologies proliferate, gateways need to evolve. The best smart gateways are designed to support several use cases across various domains, such as smart homes, energy and industrial scenarios, and require REST API-based SDKs for application development. Additionally, the gateways can support all major interfaces and can handle software upgrades, semiconductor platforms and a wide range of devices.
Gateways perform crucial edge analytic functions. This is a major factor behind their rapid market adoption. They analyze data coming from a device before it is sent to the cloud, meaning analytics are performed faster and without the need for vast amounts of storage and processing power. Programmable edge analytics help implement innovative use cases in a short timeframe, and are particularly attractive in situations where autonomous decision-making is required for critical operations.
One of the main benefits of the IoT gateway is added security. Gateways both protect data in the cloud or travelling to the cloud, as well as ensure that external unauthorized parties do not attack or take over IoT devices. Gateways help with compliance, too, a huge benefit given the complex privacy and data governance issues surrounding types and location of data. They are configurable to comply with standards (such as those defined by oneM2M) and to support proprietary interfaces.
Improved time to market
Gateways have an important role to play in product realization. Gateways with pre-integrated protocol interfaces and use case scenarios, such as smart metering or smart parking, can reduce time to market by four to six months.
There are other ways that the new breed of gateways can speed things up. Modern gateway architecture enables organizations to add new protocol interfaces much more quickly. Portability across multiple OS and pre-integrations with SoC platforms also speeds product realization and, what is more, there is a significant reduction in total cost of ownership when deployed across multiple product lines.
Great… but is there a downside?
The truth of the matter is that IoT gateways have a lot going for them and are here to stay. However, edge analytics are a double-edged sword. Yes, great efficiencies are gained in terms of speed and processing power, but what is sacrificed is a lot of data, effectively “left behind” as the “relevant” data heads to the cloud. That’s great for many situations, but not all.
Similarly, some IoT situations simply don’t need the advanced capabilities intelligent gateways offer. This is true in situations when the device has enough built-in storage, when the network has sufficient bandwidth or when the task in question does not require a great deal of processing. However, advanced processing capabilities are needed and intelligent gateways are hard to beat.
All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.