"Multiple LPWAN technologies will co-exist for the foreseeable future." - Bill Harpley, Astius Technology Speaker Q&A

Bill Harpley, Founder of Astius Technology will be speaking at the LPWAN World Forum being held on the 9th and 10th of November at Canary Wharf, London.

Astius Technology provides a range of consultancy, implementation and training services in Linux, Cloud, Wireless and Internet of Things (IoT). Bill Harpley is also the founding member of a community based IoT Network called the Brighton Everynet whose aim is to build a LoRaWAN network across the Greater Brighton region of the south coast.

Bill, who will be discussing Creating an Ecosystem to Fuel the Growth of the Industry on the 10th of November, has taken time to complete our LPWAN Q&A:

 

Which do you consider to be the best use cases for LPWAN technologies?

The broad categories of use cases which are optimum for LPWAN are those which require:

  • A long battery life (5+ years)
  • Infrequent requirement to read data (perhaps as little as once a day or once a week
  • Sensor end-devices are scattered over a large geographic area

Beyond that, I think you are looking at commercial considerations such as ROI and market size. For example, a scenario which called for the dense deployment of sensors could meet both technical and commercial criteria for success. (e.g. Smart Cities, Smart Meters). You could think of these as “mass market” deployment scenarios.

Another set of interesting use cases are those which serve a specialised niche. For example, the Things Network in Reading was recently used in a project for tracking hedgehogs. This is not a mass market solution but it does meet a very specific need.

 

Is there a place for multiple LPWAN technologies to co-exist long-term or will there eventually be one dominant technology?

I expect that multiple LPWAN technologies will co-exist for the foreseeable future. 

We already see this situation with short-range wireless technologies. System designers can choose from Bluetooth, Zigbee, WiFi and a host of others.

It really is a question of matching the technology to the application: battery life, security, range and lifetime cost are some of the factors that need to be considered. So given that there is such a broad range of IoT application requirements, it may not even be desirable to think about a single dominant technology.

 

How can LPWAN security concerns be overcome?

I think the problem should be more broadly framed in terms of end-to-end IoT security. This is a tricky issue, as the IoT technology stack will be built from components and services which are supplied by a variety of vendors. A security flaw in any one of these components could compromise the whole stack.

Referring specifically to LPWAN infrastructure, we need to think in terms of using:

  • Mutual authentication of communicating devices (e.g. endpoints and gateways), otherwise there is the possibility of ‘rogue’ endpoints being deployed on the network
  • Encryption of data may be desirable for many types of data (e.g. medical monitoring)
  • OTA distribution of security patches needs to be mandatory
  • In some cases the physical security of end-devices may be an issue
     (e.g. smart meters)
  • Security of end-devices needs to be considered in terms of their complete lifecycle. Basically this means that at end-of-life, devices need to be decommissioned and “scrubbed” of any data or firmware that might be useful to an attacker.
  • Perhaps extend ISO27001 Information Security Management standard to deal with these types of issues.

Perhaps the biggest current problem is the default state of end-devices, particularly for consumer products. These often come with default passwords, primitive security and no facility for OTA updates.  This situation can only be fixed by charging more for these types of products but whether consumers would be willing to pay extra remains to be seen.

What is certain is that LPWAN deployments potentially represent a massive increase in the available attack surface for malicious actors. So ultimately, it may require primary legislation to prescribe security standards.

 

How best should telcos position themselves in the LPWAN ecosystem?

This is perhaps the most interesting question of all.

Telcos will only succeed in this game by both exerting influence and being a team player. The object of the exercise should be to help build a market for LPWAN services and demonstrate to potential customers that there is real business value in subscribing to these services.

Listen to what customers want. In some countries you will see some operators offer both cellular and non-cellular LPWAN solutions. Indeed, this is already happening in many countries such as South Korea and the Netherlands. This will allow operators can offer LPWAN services at a range of different price points.

By contrast, dominant market players such as Vodafone plan to offer only NB-IoT services.  This may be a feasible strategy in countries like the UK, which have no national LoRa or Sigfox network.  However, I am not sure that this is a viable long term strategy. It may simply open the way for an independent operator to rollout a national network based on LoRa or SigFox.

Partner with verticals. Telcos who hope to make big bucks from LPWAN revenue streams may be disappointed. At best, any increase in revenues may only offset the decline in revenue from voice services. So a better strategy would be to form partnerships with large vertical players in transport, energy, environmental services, etc. That would allow them to get a slice of the revenue from applications.

Influence & Enable others. Play an active role in standards bodies and other industry forums. Collaborate with downstream suppliers. Establish developer programmes.

 

How disruptive will the commercial deployment of NB-IoT be?

NB-IoT will have a significant impact, though I suspect not as much as its advocates hope.

There is a sense in which in order to succeed, cellular operators  must first disrupt themselves.  This is not as simple as it may sound:

  • Operators in some countries are planning to ‘sunset’ their 2G and 3G networks within the next few years. By contrast, in the UK we can expect to see these operational for perhaps another decade. For example, I have read that Vodafone have said that they will not switch off their GSM network until 2025, at the earliest. So GPRS and 3G data services could be around for quite a long time to come.
  • Many customers are on long-term contracts based on legacy technologies. When these contracts end, operators will seek to migrate them over to NB-IoT technology (if the customer application is suited to this). In countries which have a substantial LoRa or SigFox presence, some customers may choose to ‘jump ship’ and take the latter path.
  • Perhaps the biggest barrier to rapid adoption of NB-IoT is the need to create a complete business eco-system. LoRaWAN already has this, as does SigFox to a lesser extent.

For these reasons, expect NB-IoT adoption to be slow and non-disruptive. Think more ‘evolution’ than ‘revolution’

 

Will the delay in NB-IoT deployment jeopardise its traction long-term?

Many countries already have market-proven deployments of non-cellular LPWAN technologies such as LoRa, SigFox and RPMA. Expect adoption of NB-IoT to be quicker in markets where these types of networks have no substantial presence (the UK being a prime example).

As noted, technologies such as LoRa and SigFox have a head start in terms of building out a viable business ecosystem. So NB-IoT has a lot of catching up to do.

For these reasons, I think that delay in NB-IoT deployment could be detrimental in the
short-term. However, it is destined to be a key plank of 5G IoT data networks, so I would expect it to be a success in the medium term.  

 

If you have been involved in an LPWAN deployment, what were the key challenges and how can they be overcome for future deployments?

So far, I have been involved in implementing a community LPWAN in Brighton. The main obstacle has been persuading the local universities and Brighton & Hove Council to provide support for the project. In the case of the universities, we wanted them to purchase and host a LoRaWAN base station. In the case of the council , we wanted to use lamp-posts to host endpoints (sensors) but were politely told that BT Openreach had taken out a 10 year franchise on the lamp-posts in Brighton. Naturally, BT Openreach declined our request for a meeting.

There are no easy answers to these types of problems. It’s a toxic mixture of apathy and aversion to risk. In general, these types of organisations are only interested if there is funding available and a major corporate brand name involved. Interestingly, this seems to be a cultural problem which is specific to the UK.  Things Network project in other regions of the world have found enthusiastic backers among local universities and city councils. Examples include  Zurich and Sydney.

 

What should be taken into consideration when deciding which technology to use for a specific use case?

In no specific order, I would say:

  • Cost
  • Security
  • Availability / Reliability
  • Coverage
  • Battery life
  • Quality of Service
  • Frequency of Message updates

Ultimately, it depends on the application. For example, monitoring street bins to see when they are full would likely have Cost and Coverage as the primary criteria. By contrast, some kind of patient home monitoring would be less price sensitive, so other factors such as Security and QoS would dominate.

 

Can market potential be achieved with different technologies and a fragmented industry?

Yes, most certainly. I think that the reasons for this are:

  • There exists a very broad range of IoT applications. Each of these will have its own price point and performance requirements for market adoption. So customers will choose the technology which is best suited to the application.
  • Overall, the market is potentially huge, with billions of connected devices forecast to be connected to the Everynet within a few years. So it is big enough to support several LPWAN technologies, especially since they each have different performance characteristics.

Although a degree of competition is good, I think that too much fragmentation would be a bad idea. So there is a lot of merit in proponents of both cellular and non-cellular technologies engaging in dialogue and developing a shared sense of mission for the industry. That is why I think that events such as the LPWAN World Forum are so valuable.

 

What does the future hold for LPWAN?

  • We will see further progress in energy harvesting, to further reduce dependence on batteries (or eliminate the need for batteries altogether)
  • Spectrum will continue to be in short supply. Some sharing of spectrum between licensed and unlicensed LPWAN may be feasible for certain types of applications. We can already see a similar arrangement with LTE-U, which provides for operation of LTE in unlicensed Wi-Fi spectrum bands.
  • We need to drive the cost of collecting data towards zero, so expect continued pressure on margins for suppliers of hardware and network infrastructure.
  • The people who will make the most money will be the people who own the data.
  • Both cellular and non-cellular LPWAN technologies will play an important role in the 5G technology ecosystem.
  • Unsuccessful technologies will fade from view. We have already seen one example of this: products based on the Weightless family of standards have gained only limited market success. So already, we can see a Darwinian process at work.
  • Focus will shift away from LPWAN infrastructure to the creation of new types of products, services and user experiences, based on the data collected from the network.

 

For more information please visit: http://lpwanforum.com/

 

About LPWAN World Forum

LPWAN World Forum is a 2-day conference and exhibition covering Low Power Wide Area Networks and its impact on IoT technology. Hear from over 40 visionary speakers and leading case studies on how LPWANs are enabling a much wider range of M2M and Internet of Things (IoT) applications. Take part in roundtable discussions for in-depth learning and network with peers and colleagues in our exhibition and demo area. Gain a unique insight from the major industry contributors. Make sure you are present in London on the 9th and 10th November for THE IoT Networks event of 2016.

www.lpwanforum.com