More
    HomeInsightsBT backs LoRa and NB-IoT for smart cities, eyes future LTE-M trials

    BT backs LoRa and NB-IoT for smart cities, eyes future LTE-M trials

    -

    BT expects a “horses for courses” approach to IoT connectivity with the operator opting for LoRa, NB-IoT and LTE-M in its smart city deployments and lab research.

    Speaking to Mobile Europe on the first day of its Innovation Week, Howard Watson, CEO of BT’s Technology, Services and Operations unit, said LoRa has already been deployed in Milton Keynes and London.

    The MK-Smart project in Milton Keynes launched in late 2016, with BT working with the Open University and the city’s council.

    BT’s MD of Research and Innovation Tim Whitley said three major applications are active: smart refuse collection, smart parking and pest control.

    As for NB-IoT, Whitley told Mobile Europe that it will be deployed in coming months, although he declined to reveal a firm launch date or locations for the network.

    While the operator is currently focusing on these two technologies, Watson said the other major 3GPP-standardised IoT technology, LTE-M, is also currently being explored by BT in a lab context.

    While it is not currently being tested in the field, this is “on the roadmap,” Watson said.

    Explaining the decision to initially focus on LoRa and NB-IoT, Watson highlighted the low cost of both as a key selling point.

    He said: “The good thing about either LoRa or NB-IoT is that they’re actually relatively inexpensive to roll out because you don’t need large amounts of backhaul.”

    BT is far from the only operator to be exploring a number of IoT connectivity technologies. France’s Orange is using a combination of LoRa and LTE-M, while Tele2 is using a similar NB-IoT and LoRa approach to BT.

    According to BT’s Whitley, both technologies also fit the applications that BT has been looking at in Milton Keynes.

    He said: “My view is both these technologies will coexist for a while because it’s a very broad application layer, because we’re looking at the totality of everything we do and how we can make it a bit more efficient by collecting tiny bits of data from devices, people and sensors.

    “The one characteristic you get from all those applications is that they are generally more telemetry-type data, with very low data rates.”

    He said battery lives rather than latencies, for example, were more important in these kind of deployments. He highlighted meter readings as an application where missing the occasional transmission does not matter.

    “You can trade certain performance there. There are other types of IoT where it’s really important that you have guarantees about reliability.”

    The thinking behind the projects, according to Whitley, is to provide the connectivity and then harvest the information from it to allow customers and businesses to do things “that bit better”.

    One example of this data harvesting is the MK Data Hub. According to Whitley, this is a layer that allows networks of IoT devices to easily deposit information.

    Data from across the city can then be collected and curated by a central hub in order to apply policies such as rating. Developers or others can subsequently use it to build applications. Whitley calls this “democratising” data.

    “The way I describe that element of IoT is it’s almost like task-omniscience. If you have a task, what do you need to know in real time to do it? Really, IoT and LoRa is about this.”

    Whitley said he thinks there will be space [for multiple technologies] “for a while as we work out the value chains and the winning use-cases”.

    He added: “It’s going to be a bit horses for courses.”

    In the longer term, though, Whitley is enthusiastic about the role that 5G will play in enabling IoT connectivity.

    This is not just in improving performance, but in the potential to provide dedicated virtual networks that offer different qualities of performance to different applications with different requirements, or 5G network slicing as it is more concisely known.

    This means “ultimately about making sure you have the right performance for the right application”, he said.