3 key questions for winemakers wanting to transform their business with the Internet of Things.
IoT, Smart Farming, Precision Agriculture. We've heard these terms but how do you successfully implement it?
The Internet of Things, IoT, Connected Vineyard, Industry 4.0, Smart Farming, Precision Agriculture.
It’s likely that these are terms you’ve seen and heard for some time, but how do you actually implement these concepts successfully?
What we’re fundamentally talking about here with the Internet of Things (IoT) is making use of connected sensors to allow us to make better informed and data-driven decisions on what do in the vineyard and winery, and to learn from that data.
It also allows us to see what’s happening in places that you typically can’t see easily and measure things that are typically hard to measure.
On first sight it all seems pretty easy but the reality can be quite different.
Here’s some questions you need to be asking, of yourself, partners, and suppliers, before even thinking about implementing any of this technology in your vineyard or winery.
Data
Data is everything.
This is by far the most critical factor that can make an IoT investment a success or a failure. We need to know what data to collect, from where, how often, and importantly why we’re collecting that data.
Many IoT projects fail because they simply try to collect as much data as possible from as many places as possible, without really thinking about why. Similarly many fail because they simply collect data that they were already collecting they now just automate the collection, they’re not really adding anything new to their decision making.
We have to understand what data we need, and where it’s going to come from but we also have to understand how often we need that data. Does it need to be real-time, is once per day enough, do we only need it at certain points in the process?
Finally on collecting data we need to know how precise and how accurate it needs to be. If we’re measuring temperature, for example, does it need to be to 0.01℃ or is to 0.1℃ ok? What if the sensor has an error of +/- 0.1℃? Is that acceptable?
Now that we’ve got a load of data we have to know how we are going to analyse it.
Does it need to feed in to some mathematical equations, some winery management software, are we trying to do some learning on previous data and need some data science tools? How are we going to visualise the data, do we need to create some dashboards with charts and various statuses? What software are we going to use do that? What we need to understand here is what we’re going to with the data once it’s collected.
Key question: What questions am I trying to answer with my data?
Connectivity
Connectivity for IoT is not quite as simple as plugging in a Wi-Fi router at home.
There’s a huge range of options to suit differing requirements.
Your requirements, and some constraints such as budget and location will determine which of the huge range of connectivity options are best for you.
For example there’s relatively cheap options such as LoRaWAN which is a standard specifically designed for battery powered sensors with low data requirements and spread over a long range, hence, it’s in the LPWAN (Low Power, Wide Area Network) category.
Something like this will give great coverage, potentially over Kilometres, but speeds are in the low Kbps so you’re definitely not going to be getting video feeds from this, but a field of temperature sensors would work just fine. Someone with good tech skills could likely set this up themselves on a small scale and there’s even existing community-run public networks you can use to test ideas before going in deeper.
On the other end of the scale are options such as Private 5G Networks. There’s been significant interest in using 5G for IoT networks with factories, ports and even farms (yes, smart Cows are a now a thing) testing and building them.
As the name suggests they use the same 5G technology that your mobile phone uses but are built for private users. So for example, you could build a Private 5G Network that covers your vineyard and winery, it would be completely owned and controlled by yourself and only your devices could connect to it, unlike a Public 5G Network that the general public use. These networks are more expensive to build but have outstanding features such as wide coverage, high speeds and mobility (i.e. devices can easily move around with connectivity interruptions). Building a Private 5G Network requires specialist skills and equipment meaning you’d need to hire a specialist company to help you.
Understanding your requirements around all of these aspects will help you in deciding which connectivity method(s) is best for your situation now and in the future.
We need to be thinking about: coverage, speed, budget for set-up and running, spectrum availability, and what we want to connect.
Key question: What are my connectivity requirements?
Security
Security is something that people often forget or simply don’t care about, but it’s one of the more important aspects to consider.
The data being moved across IoT networks could be critical to your business and so making sure its secure is also critical.
There’s two main areas to think about on data security: data in transit, and data at rest. Data in transit being the data while it’s being moved between servers and sensors and very likely off in to the internet in to some cloud-based software. While data at rest is the data that is stored in servers and used by software once it’s been transmitted.
We should therefore be thinking about security in every aspect of our IoT deployment, from sensors on the ground or in the winery through to the cloud-based software we use to visualise the data.
When we think about Security we should be thinking about risk.
What is the risk of data being stolen? How likely is this to happen and what are the consequences if it is?
What is the risk of data being manipulated? How likely is this to happen and what are the consequences if it is?
For example, if someone wanted to damage your business they could breach your IoT network and simply change some of the data to give incorrect readings. What impact would this have?
It may be minimal but may be more severe if, say, your irrigation was automated and you ended up over watering your vines because the data kept saying that soil moisture was too low. While the likelihood is low the impact may be severe.
There’s a wide range of security threats to consider in this same vein, from Denial of Service through to Ransomware.
By understanding these risks we can understand our own risk appetites and implement solutions that satisfy that risk appetite. This avoids us having little to no security but also avoids us going over the top and splurging our entire budget on military-grade security which is unnecessary.
Key question: What level of security do I need for my data?
BONUS: Automation
With a well functioning IoT system we get lots of data and can make more data-driven decisions which can transform a business.
Following this comes an opportunity to automate. High quality data allows us to make high quality decisions, but does it need to be us humans making those decisions on a day-to-day basis, possibly not.
By thinking about what in our vineyard and winery could be automated if high quality data was available we can focus our human efforts on more creative and valuable tasks that computers can’t do.
Bonus key question: What can be automated if high quality data is available?
In summary, there’s a lot to consider before implementing an IoT network in your business. This shouldn’t put you off though. Spend a bit of time thinking and documenting answers to these key questions and you’ll be far further ahead than many.