Skip to main content

Please note that u‑blox no longer supports Internet Explorer.

We recommend using the latest version of Google Chrome or Firefox.

Ok

Fulfilling the Promise of Edge Computing in the IoT

construction

While many of today’s connected tech devices take advantage of cloud computing, Internet of Things (IoT) manufacturers and application developers are starting to discover the benefits of doing more analytics closer to the devices themselves. Why? Among other reasons because the massive amounts of data being generated by IoT applications are creating a bandwidth bottleneck.

Edge computing - the ability to do advanced processing and analytics either on the device itself or close by on a local server or gateway - is one solution to that problem, Rather than streaming all of the gathered data to the cloud for analysis, thus absorbing the additional network latency this entails, edge computing takes over some of the analytics and sends only the processed information that matters through the network instead of transmitting raw streams of sensor data.

Avoiding unnecessary device‑to‑cloud data round trips ensures a faster system response and lessens the workload of the cloud, reducing the cost of building IoT infrastructure. Security and privacy can also be improved with edge computing by keeping sensitive data within the device.

Bringing intelligence to the edge of the network is especially valuable on devices that connect via wireless cellular connection such as smart meters or asset trackers. To do so, new technologies such as LTE‑M and NB‑IoT networks have been specifically designed for IoT or machine‑to‑machine (M2M) applications with lower data rates. These licensed low power wide area (LPWA) networks are brushing aside cost and other limitations that have hindered deployment of extensive wireless sensor networks. In other words, the more you can compute locally, the less data you have to send over the network. Overtime, this would offset the fact that you do need to add a processor to your device to do all this computing in the first place.

Applications such as crop monitoring, livestock monitoring, and other asset tracking apps will be among the first to benefit from these technologies. Let’s look at these networks one at a time:

Narrowband Internet of Things (NB‑IoT) is a cellular technology specializing in communication between “things” that require small amounts of data over long periods of time in hard to reach places such as long distances from a cellular base station, or in shielded areas such as deep within buildings or underground structures.

Developed by 3GPP, which unites seven telecommunications standard development organizations, NB‑IoT enables IoT connectivity with extended coverage, lower power consumption and the security of cellular for a wide range of devices and services.

What is more, the existing cellular standards don’t support the level of power saving capabilities that NB IoT does, which makes these standards unsuitable for inexpensive devices that require battery lives of several years.

NB‑IoTNB IoT can help local government control street lighting, determine when waste bins need emptying, identify free parking spaces, monitor environmental conditions, and survey road conditions. It also provides a secure and low cost way to track assets when continuous tracking is not necessary, NB IoT connectivity offers farmers and municipalities the ability to capture data from an environmental sensor containing an NB IoT module monitoring the temperature and humidity of soil as well as the levels of water or air pollutants or the amount of rainfall, issuing an alert if anything is outside of established norms.

LTE‑M, the other licensed LPWA technology, is tailored to the needs of IoT and M2M applications that operate on a tight power budget with low to medium data rate requirements. It enables longer battery lifecycles and greater range as compared to standard cellular technologies such as 2G, 3G or LTE.

LTE‑M can deliver remote firmware updates over the air (FOTA) within reasonable timeframes, making it well suited for applications running on devices that may be deployed in the field for extended periods of time.


Battery life of up to 10 years on a single charge in some use cases also contributes to lower maintenance costs for deployed devices where end devices may not be connected directly to the power grid.

LTE‑MLTE‑M supports full hand over between network cells from a moving vehicle and is therefore well suited for mobile use cases such as vehicle tracking, asset tracking, telematics, and fleet management. Within smart cities, LTE‑M can meet a variety of needs and effectively control street lighting, determine waste management pickup schedules, identify free parking spaces, monitor environmental conditions, and survey the condition of roads.

In summary, for reasons including bandwidth, cost, speed, battery lifetime and security, a faster, cheaper and smarter approach is needed than the traditional way of gathering data and sending it through networks to the cloud. Low power wide area (LPWA) networks and edge intelligence promises to provide a viable solution, and in this way extend the range of applications for the IoT.