The Internet of Things (IoT) or the Internet of Everything (IoE) is the flavour of the month and with it generates a lot of hype around what it will do for you and how it can revolutionise your industry. We are reading articles about how IoT will help farmers finally realise true sustainable farming or how IoT will be the catalyst that sparks the next industrial revolution. This is exciting, and the market is suddenly flooded with cool and exciting “connected products” that can help us be smarter about the way we go about processes. The problem is these devices are not really connected they need to be connected and here lies the biggest barrier to anyone’s entry into the IoT world how to solve the great connectivity disconnect.

In my line of work, I get to be involved in the IoT Solution from design to implementation and one thing rings true no matter what vertical I’m working without suitable connectivity the IoT dream cannot be realised. My approach to any IoE project is how do I implement a robust flexible communications infrastructure that can handle the scale of an IoT rollout? To understand this, we need to have a closer look at what connectivity standards are out there and what their relative strengths and weaknesses are. Armed with this information we can then design a communications network that will be able to handle the thirst for data your client will have once they discover what having access to it can do for them.

Cellular

The 3G/4G cellular network is the easiest way to get connectivity to your devices it is already available, if you are lucky enough to have network coverage you don’t require to install any additional infrastructure to access it. The issue with using this medium is that it requires a per device connection and a monthly subscription with ongoing costs. Telstra in Australia are one of the first to deploy both Narrowband coverage and Cat M1 IoT technologies to provide plans specifically tailored to IoT devices which only require to send small volumes of data at very low data rates. The Cat M1 network operates over Telstra’s 4GX network on the 700MHz frequency allocation meaning if you can get Telstra 4GX coverage you can access their IoT plans and pricing. If you are just on the edge of the network coverage there are solutions out there that can help boost your signal [Link Here to Case Study] I recommend engaging an RF Specialist to design a cellular boost & coverage system suitable for your requirements.

The Pros

• Instant access without the need for additional infrastructure
• Availability of NBIoT & Cat M1 plans making it more attractive for IoT applications

The Cons

• Expensive data plans with ongoing access costs
• Not suitable for large scale deployments (hundreds of sensors)

Ideal Use Case

I see the cellular network as an ideal connection medium for my Proof of Concept (POC) installations where I need to install one to two sensors simply to get data flowing back to the cloud, so I can prove to a client what insights are possible from these data points. This is also ideal for one off or remote solutions like tank/bore monitoring where there is no other infrastructure available. Lastly, I would use it on mission critical applications in conjunction with other mediums where the requirement to always be connected is essential.

Wi-Fi

Any IoE project that has a larger number of sensors requires a decent internal network to support it, costing for individual connections per sensor would out way the insights gained from them. Wi-Fi utilises the 2.4Ghz and 5.8Ghz radio bands for device access and can be easily deployed to cover a small to medium sized areas. Wi-Fi standards especially the newer ones lend themselves to high speed access and data transfer allowing for devices like smart phones and tablets to easily connect and access IoT control platforms. Not all wireless access points (AP’s) were created equal and deploying a reliable wireless network in a noisy and high demand location isn’t as simple as slapping up a few Ubiquiti AP’s. There are many factors which effect a wireless networks functional operation like frequency interference, structural interference, machine noise interference, antenna type, deployment locations just to name a few. I recommend seeking the advice of an RF Specialist to design a system suitable for your requirements.

The Pros

• Relatively inexpensive medium to deploy with no ongoing costs (unless you subscribe to an AP cloud management service)
• Ideal to deploy for small to medium sized areas for coverage
• Supported by a worldwide standard making it easier to implement with devices that have Wi-Fi capabilities already built in.
• Large bandwidth capacity allowing for data heavy devices

The Cons

• Highly susceptible to interference and not suitable for all environments
• Unable to handle large amounts of simultaneous connections (Clients)
• Connection coverage is limited to a smaller area typically limited to the capabilities of the end device
• Requires additional infrastructure to be installed

Ideal Use Case

Due to the ability of Wi-Fi to provide large bandwidth it lends itself to data or bandwidth hungry devices like Smart Phones & Tablets allowing them to easily access rich content portals like MyInsights Cloud Platform. This type of network is also ideal for CCTV cameras as it can keep up with the demand of streaming HD Footage when required. Keep in mind when choosing this type of medium as your communications bearer it will require some form of gateway to access the internet. I would also suggest keeping your IoT infrastructure separate to your company’s operational networks due to the amount of traffic an IoE solution can generate.

Zigbee

Anyone operating any of the older M2M type devices would be familiar with this standard and it is still being rolled out as a connection standard by much of the big players in the industrial sensor/PLC SCADA control market. Zigbee is a wireless protocol that uses low power, low data rates and requires to be in close proximity of the gateway device. Due to its low power consumption its distance is limited and requires line of sight. Zigbee can overcome distances by passing traffic through a mesh network of intermediate devices to reach ones further out in the network.

The Pros

• Great for applications where wireless is the only solution and the distance to cover is less than 10m
• Devices ability to mesh can help to extend the networks reach

The Cons

• Whilst Zigbee is an open source protocol most manufacturers that employ this system use a proprietary version of it which locks you down to using only their products
• Connection distance is very limited and requires you to purchase more sensors to extend it
• Usually in the industrial space these can be very expensive systems to deploy and are not very scalable for large areas

Ideal Use Case

If you are looking to deploy an ecosystem from solely the one supplier and are happy to fork out significant amounts of money to do so, then this maybe a suitable solution for your clients. I would personally stay away from this system as it significantly restricts your options in terms of sensors used and device control when it comes to the automation of processes. I think this standard has found its niche in the home automation space where the ability to mesh a network is more achievable.

SigFox

SigFox sets up antennas on towers to receive data transmissions on the 920.8Mhz Uplink and 922.3Mhz Downlink (Free unlicensed ISM band in ANZ). SigFox sends very small amounts of data and at very slow rate because of these small packets sent at 300 baud it can achieve transmission at very long distances. To access the SigFox network you need to pay a supplier for access with an ongoing monthly subscription type model like the cellular network operators.

The Pros

• Transmission over long distances
•  Low power consumption enables longer battery life
• No need to install additional infrastructure

The Cons

• Paid subscription type model requires ongoing costs (from around $2/device per year)
• Coverage areas in Australia is poor and limited to small pockets
• Not suitable for real time data applications
• Proprietary protocol
• Not suitable for high bandwidth devices

Ideal Use Case

SigFox is ideally suited for solutions where a device only needs to send small amounts of data and at irregular intervals, something like asset tracking is an ideal use case. This obviously depends on your intended use areas have SigFox coverage, I would consult with an RF Specialist to design the system around its intended use case to ensure you will get the connectivity you require.

LoRa

LoRa is another low power small data transmission protocol like SigFox, where it requires a base station deployed on a tower to receive data transmissions. Where they differ is that LoRa is an open source protocol meaning you can deploy your own base stations and gateways to add LoRa coverage to your property. LoRa has transmission distances of up to 25km and a base station can handle up to 1000 nodes connected to it.

The Pros

• Transmission over long distances
• Low power consumption enables longer battery life
• Open source protocol
• Base Stations can handle up to 1000 nodes connected to it

The Cons

• To use this bearer, you will need to either subscribe to a provider or install your own network so additional infrastructure is required
• Not suitable for high bandwidth applications
• Not a lot of viable solution options at present but this market is growing

Ideal Use Case

LoRa is the perfect solution for a client that needs to deploy hundreds of sensors like soil moisture probes that only require to transmit small data payloads but could be spread over a large area. This is also perfect for a rural solution where one base station could cover several properties and they could all access the same base station reducing the cost for entry into the connected space. Deployment of these networks is not as easy as it sounds I have received a lot of feedback from companies trying to deploy this technology and the difficulties they have faced in doing so. I recommend seeking the advice of an RF Specialist to design a system suitable for your requirements.

As you can see solving the connectivity barrier can be quite a challenge and needs to be treated on a case by case basis what will work for one client may not be suitable for the next. My approach when it comes to the complete solution is to combine a number of these communication bearers to achieve a solution suitable to all the client’s business applications. To realise the dream that is Smart Farming or the next industrial revolution we must design robust reliable and scalable connectivity solutions that ensures continued real time access to IoT data.