Semtech is a US-based supplier of high performance semiconductors and advanced algorithms for communications and industrial end-markets. The company’s LoRa devices and wireless radio frequency technology – LoRa – offers long range, low power wireless connectivity that has become a leader in the low power wide area (LPWA) networking market worldwide. The company is a founding member of the LoRa Alliance and a member of the IoT M2M Council (IMC). Here, Therese Cory, a senior analyst at Beecham Research, interviews Alistair Fulton, the vice president of IoT product management and marketing at Semtech.
Therese Cory: What do you see as the role of LoRaWAN in enabling the IoT?
Alistair Fulton: The wireless remote monitoring of assets relies on understanding the condition of these assets, through the transmission of data from IoT-enabled devices to IT systems for analysis. Most applications of this type do not need deep analytics, and data aggregation and display is sufficient. Semtech views LoRa as the best wireless technology available today for industrial use cases. The utilities and water industries, where small amounts of data are sent intermittently, and low power devices can work buried in the ground for many years without being dug up, are particularly well-suited.
TC: Do you see LoRaWAN as sufficiently accurate and low power to enable the tracking of the billions of smaller assets which ultimately will make up the IoT?
AF: Low power and accurate geolocation provision afford the best combination for tracking the small assets that will ultimately make up the IoT. The value of many IoT solutions is understanding where an asset is, as well as its condition – including temperature, humidity and other data. That kind of vision requires the availability of low cost and easy to implement connectivity. Using GPS combined with cellular techniques provides sufficiently accurate location data for connected assets today. However, looking to the future where billions of assets will be connected, customers will have to think carefully about the scalability and affordability of building IoT deployments on this scale. If one wanted to determine the location of every one of these assets, the cost would be prohibitive with today’s technologies.
For example, taking a sampling approach, measuring location for one in every 100 devices then using extrapolation to estimate the other 99% would be barely affordable today; we estimate the costs involved relating to the components and their power usage would really limit implementors to examining only one in 100 devices or fewer. We believe that compared with other wireless technology, LoRaWAN can rise to this challenge by affording the scalability needed at affordable cost.
TC: Why is developing IoT solutions too complex, time consuming and costly? How is Semtech helping developers and solution providers for this purpose?
AF: Building IoT solutions with today’s resources and skills necessitates developers to have deep knowledge of RF technologies and expertise in embedded systems, particularly in a brownfield scenario. For customers buying an IoT solution, cost is the definitive issue, particularly for a solution that must be scalable.
Semtech aims to assist both developers and solution buyers by making IoT solutions both accessible and low cost, by supplying a cheap, malleable set of development tools based on LoRa. These include a cloud-based tool which works by extracting key network and signalling data, sending them to the cloud where a specially designed algorithm is applied, and where the resulting geolocation info can be retrieved through an application programming interface (API) call. In early 2018, the company began trialling a simple cloud-based service for the purpose; some 500 customers are using the service, sending several million requests a day – though this is still relatively small scale compared with a full implementation. We expect commercial service to be available in 2019.
In addition, Semtech along with the LoRa Alliance are helping implementors through offering app templates and simple building blocks, as well as offering graduate level training resources to help developers work with LoRa’s idiosyncrasies. They are working towards a full end to end reference solution for common use cases such as asset tracking.
