This article presents a roadmap that guides companies to implement IoT in their organization, creating value and a future-proof company. Internet of Things is often misunderstood for only consisting of sensors and a network. However, as discussed in the first article, IoT only adds real value when incorporating the new information in business processes. Deloitte University, shown in Figure 1, captures this idea in their information value loop. Each stage is preceded by a technology, together forming a complete IoT implementation. For each stage, I will discuss the roadmap and common pitfalls.
Stage 0: The foundation for success
A company should not dive into the creation stage straight away, because these efforts usually end up in a drawer. Critical for success is a business case to verify added value, a technological concept to verify feasibility, and a common sense of direction.
- The business case should be based on KPIs that define value for your organization. When aimed at internal improvement, the case should preferably go beyond simple detection and control towards transformation of business process, as this is where most value lies (McKinsey, 2015). When aimed at improving products and service, the case should focus on adding IoT capabilities to existing products and services (McKinsey, 2019).
The technological concept defines how it will be done, for each layer of the technology stack: devices, connectivity, platform, and application development. Bhalekar and Eloot (2018) recommend defining a core architecture choice. The advantages of one architecture is reusability and interoperability for all IoT cases within the company. This architecture should enable scalability for individual IoT cases and for integrating the different IoT cases. Evertson (2015) classified various architectural styles for IoT, which is a good starting point for defining the core architecture.
A common direction is necessary, because of the variety of domains involved: IT, management, operations, and everyone involved in the business process or product that will be transformed. McKinsey (2019) recommend three ways through which this common sense of direction can be achieved.
- Defining how IoT will create value, which is strongly linked to creating a good business case.
- Take initiative from a senior leadership level. This shows commitment and helps overcoming barriers.
- Involving all departments in the IoT creation and implementation process. It is the nature of IoT that it crosses departmental borders, thus requires alignment across functions within a company. Especially when pursuing an IoT strategy, it involves developing products, improving services, and optimizing processes.
Stage 1: Create
The first stage aims to use sensors to generate information about a physical event or state. Sensors are mature technologies, widely available at a low cost. Design of hardware often requires custom engineering, tailored to the use case. Especially for companies with specific objects, such as products and machinery. Partnering with an engineering company is recommended, as these skills are not easily developed in-house nor widely available. For widely used equipment, such as storage systems, complete sensor solutions already exist. In those cases, it is recommended to find a vendor. A middle ground are vendors that sell complete and generic hardware, such as VersaSense, which can be used in a variety of contexts.
Stage 2: Communicate
Communicating involves the transmission of information from one place to another by means of a network. A large variety of networks exist, from long-range networks (e.g. LPWAN) to short-range networks (e.g. Bluetooth). The choice of network depends on the use case. The most important criteria are:
- Power consumption, from 10 years on one battery to being connected to a power outlet.
- Device computation power, depending on physical constraints.
- Communication data size, depending on what data is collected by the sensor.
- Communication reach, from a single room to worldwide.
- Security requirements, such as encryption and key management.
The core architecture defined in stage 0 has to enable the usage of the desired networks, which is part of the next stage: aggregate.
Stage 3: Aggregate
Aggregating means gathering together information created at different times or from different sources. IT infrastructure is required to be able to aggregate. This is a challenge, as a large variety of sources are used in IoT. Companies often use specific IoT platforms to address this issue. According to Lazarescu (2017), these commercial solutions are effective in addressing vertical application domains. However, these IoT platforms can lead to vendor lock-ins, which can hamper business potential for further IoT cases and horizontal integration. Therefore, a more generalized solution that can integrate everything, such as eMagiz, is recommended.
Stage 4: Analyze
Analyze is the discernment of patterns or relationships among phenomena that leads to descriptions, predictions, or prescriptions for action. Analytical tools, also called augmented intelligence, are required to enable these analyses. These technologies are increasing in numbers quickly, as more and more data is becoming available. Various companies specialize in specific types of analysis, each providing their own software, ranging from generic inventory optimization to specialized predictive maintenance.
Bhalekar and Eloot (2018) recommend incorporating data scientists up front. They can advise on which data has to be collected, which analyses can be made, and help deciding the software to use. They can also collaborate with internal business process specialists on how and which data should be collected, to ensure accuracy and relevancy.
Stage 5: Act
To complete the cycle, action needs to be taking: initiate, maintain or change a physical event or state. Augmented behaviour are technologies that improve compliance with prescribed actions. As discussed in the first step, most value can be created by transforming business processes. An example is the transformation of vendor-managed inventory process, shown in Figure 3, which was my Industrial Engineering thesis research. The traditional process is linear, where customer and vendor need to wait on each other, creating a slow and inefficient process. The IoT-based process does not require any action from the customer, creating much faster process. This example shows how IoT needs to end with actions, which can be realized by integrating IoT in existing business processes and apps. Here, the inventory management system (SCEM) made replenishment orders based on customer input. In the new situation, replenishment orders are based on IoT input.
Now the information value cycle is complete. Built on a solid foundation: a business case, technological concept, and common direction. Creating sensors, making them communicate, aggregating their information, analyzing the data, and acting upon these insights. That is how IoT is created and implemented. The result? A competitive advantage that makes a company future-proof.