For over two decades, people around the world have been working with IoT[1]. The first decade was primarily driven by Proof of Concepts, exploring how to connect more or less silly devices to interact with the outside world. Moore's Law has caused devices to shrink in size and cost, and with today's affordable and battery-friendly sensors, the possibilities with IoT are growing to infinity.
Now that IoT has really come out of the testing stage and is increasingly being considered in business-critical contexts, in this article I want to highlight three key focus areas that you should be aware of when working with IoT as part of a digitalization project.
1. It doesn't start with the gadgets, it starts with the business
Between 2018 and 2023, the number of devices connected to the internet will double[2]. It can be tempting to simply follow the trend in your digitization project and throw yourself into installing various sensors and then investigate what you can get out of the data that is collected. IoT efforts driven from the IT department of a company will, in its most caricatured form, be driven in this way.
It's not surprising that the digitization project must be driven by the business needs. The first significant barrier[3] is getting buy-in from top management. One step to overcome this barrier is to articulate IoT as 'just' one of the many tools in the enterprise toolbox to gain deeper insight into data about the business. Just as it would be to look at data from the financial system or inventory management system if you are a manufacturing company. What's new with IoT is that you have the opportunity to look at data from points in the company that it has not previously been possible to get data from, or at least not such detailed data. In other words, IoT is helping to move towards a more data-driven business.
The first thing to ask yourself and write down in your business case is: What are the business benefits we want to achieve - why do we want to run the project?
Examples of why:
- Streamlining - optimization
- Better utilization of assets
- Improvement of service experience
- New perceived service
It's important to stick to your 'why' and let it permeate the project from start to finish. Because just like in other digitization projects, there will most likely come a time when the project is under pressure to deliver, or when the technical possibilities slip from the original business needs. It can be tempting to focus on delivering a technical deliverable and show that you can deliver an IoT project as promised, but this can result in an unrealistic deliverable and provide a fragile foundation for the journey ahead. One approach to continuously get feedback from the business is to use an agile approach, such as Minimum Viable Product[4] (MVP). This way, you can continuously pressure test the data collection against the business needs, and it may well be that experience means that the 'why' you are working with changes throughout the project, as you become wiser based on data collection.
2. There must be a balance of efforts
As with many other modernization initiatives, IoT works with a maturity scale. Loosely inspired by Porter et al [5], there are four stages in the IoT maturity model:
- Monitoring. You are able to monitor with sensors and collect data that you can then analyze and follow up on.
- Control. You make the system react to the monitored data, for example, you can turn off the heating when the desired temperature is reached.
- Optimization. With algorithms, your IoT system will be able to continuously optimize its performance or, for example, tell you when it's time to perform maintenance on a production machine, also known as predictive maintenance in Industry 4.0
- Autonomy. Enable your IoT solution to learn and make decisions for further optimization based on monitoring, control and optimization. This is where Artificial Intelligence (AI), Machine Learning, etc. come in and work with data captured in the physical world.
Nirvana is not always at maturity level 4. There must be a balance in the investments you make in your IoT efforts in relation to why you have initiated the effort. It all starts with the business case in relation to whether you should develop a solution at level 1, 2, 3 or 4, and whether you should launch an IoT solution at all. For example, it doesn't necessarily make sense (financially) to create a water waste monitoring solution for DKK 1,000 per year if you can save at most DKK 800 of water per year. Unless, of course, you have a level-4 solution that turns off the water when you've had your share of shower time. In some cases, however, it's not always possible to see if there are opportunities in the future if you go with a higher maturity level than necessary. Here, you can make a conscious choice to go with an over-engineered solution, with the expectation of reaping the potential at a later date. But not every time, and not at all costs.
3. When the project is over and the consultant has left...
It's important to realize that while an IoT initiative can start as an MVP, you need to consider scaling and embedding from the start. And an IoT initiative affects a larger part of the organization than you might think. There will typically be a need for feedback loops all the way to the employees who work close to and with the sensors when the solution has to live in the real world. Often the really good input comes from the employees who are close to operations and whose everyday lives are most affected by the IoT project.
It is also essential to integrate IoT efforts more broadly into the company's IT landscape. In some cases, it may make sense to combine data from individual sensors with data from other systems or sensors[6]. It is therefore important to look up from the individual sensor and persistently look at whether additional data can support getting closer to your why. Therefore, you need to have clear guidelines for how the IoT solution can/should be included in your digital ecosystem from the start. There are scaling considerations, of course, security considerations, including in some cases also GDPR issues that need to be addressed.
An important parameter in these considerations is whether the IoT solution performs a task with great strategic value for the company, because scaling and performance must of course be given higher priority than if the solution handles a smaller, isolated part of the company's tasks.
Conclusion
Not surprisingly, IoT projects should be approached in the same way as many other digitization projects. Three key focus areas I've covered in this article are:
- Let the business needs guide you throughout the project. Stick to the why of the project.
- Keep a balance of effort. Even though technology options are all-encompassing, choose a task that can get the job done with the least amount of effort
- Think about operations and connectivity from the start. Both in terms of how the IoT solution fits into your digital ecosystem, but especially how your operations staff will work with the IoT solution.
[1] Of course, there were IoT-like activities in the past, such as the GPS network in 1995 or when John Romkey connected his toaster to the internet in 1990, but it wasn't until 1999 that the term "Internet of Things" was coined by Kevin Ashton, MIT. 2
[2] Cisco annual report (opd. 2020): https://www.cisco.com/c/en/us/solutions/collateral/executive-perspectives/annual-internet-report/white-paper-c11-741490.html
[3] Dansk Industri m.fl. “Every. Thing. Connected.” Rapport om internet of things i danske virksomheder: https://www2.deloitte.com/dk/da/pages/strategy/articles/iot-danske-virksomheder.html
[4] http://www.syncdev.com/minimum-viable-product/
[5] Michael E. Porter og James E. Heppelmann: ”How Smart, Connected Products Are Transforming Competition”
[6] Example; Projekt fra Kalundborg påpeger yderligere potentiale ved samkøring af data fra andre systemer: https://videncenter.kl.dk/media/26199/kombit-iot-erfaringsopsamling-10.pdf
