Key Factors for a Realistic Internet of Things (IoT)

Arpan Pal, Balamuralidhar Purushothaman

This article is taken from Arpan Pal, Balamuralidhar Purushothaman book IOT Technical Challenges and Solutions and is reproduced here with permission from Artech House.

In the midst of excitement and hype, there are several practical deployments of IoT happening today delivering realistic business benefits. Examples include fleet monitoring, building monitoring, customer intelligence in retail stores, telemedicine, navigation, and tracking. The potential user community is becoming aware of this technology and what it can offer. Some restaurant chains use remote monitoring of cooler and freezer temperatures to avoid thousands of dollars in potential spoilage. Retail stores see the increased level of customer engagement that a beacon platform provides, providing customers valuable brand information and generating buzz and increased sales. There is an interesting case study on how an IoT-based energy management system delivered the return on investment (ROI) in 18 months. Smart vending allows machines to monitor their own inventory levels, calling in for deliveries only when needed and with a list of what items are needed, saving delivery drivers valuable time. Fleet and asset tracking provides not only for dispatching more efficiently but also monitoring driver behavior, which can also save on the cost of insurance and provide management with valuable information for driver training. Monitoring and processing of evaporation credits for cooling towers and landscape irrigation can save many dollars in water and sewer bills.

However, is it a great success story for all IoT applications? The truth is the count of failed IoT projects has outnumbered successful projects by a huge margin. There may be several reasons for the failures and any emerging technology may face this challenge on its travel to maturity. In this context, it may be appropriate to examine the concept of a real Internet of Things, by which we mean an IoT solution, crafted by an internal software architect or through a virtual CTO service, with an associated business model that provides ROIs on realistic business applications Even for an application focusing on social impact, it should have a business model to ensure its sustainability. For a data-gathering application, it means return of investment (ROI), the ratio of business value of the data collected, and the cost of acquiring it should be comfortably high.

There are several factors contributing to the success and failures of IoT projects and we will discuss a few of them next.

Key Contributing Factors to Real IoT

The challenges with IoT go beyond making and connecting devices that work. The integrated product and services need to work seamlessly, almost invisible to an end user. As Mark Weiser suggests in his ubiquity paradigm, we need machines that fit the human environment instead of forcing humans to enter theirs. The service needs to meet needs, easily integrate into daily life or the industrial process, and has to enhance the user’s life or the business process. This requires the reliability and robustness of all the components of the integrated solution. Sometimes the system is over engineered to deliver the application goals. The system architecture should be pruned and tuned to the business goals and deployment context.

The following are several factors contributing to a real IoT.

Robust Devices and Connectivity

Devices that work reliably in the deployment environment that do not shut down or restart due to failures or bugs are key to any IoT application. Impact and consequences of a buggy device may vary from a consumer application to an industrial application. Also, the devices should enable system level security in terms of identity, authentication, and tamper-proofing.

Another aspect to consider is the choices that one has to make in terms of technologies and standards for a cost-effective IoT system. There are many IoT standards at various maturity levels addressing various aspects of the system. One should expect to change the architecture as the technology evolves. The curse of choices for almost every component also makes the system design difficult. There are multitudes of silicon vendors, device platforms, energy sources, communication protocols, and backend software platforms from which to choose. One should also recognize that there is a trade-off between robustness and cost effectiveness, and this should be addressed within the larger context of the application.

For large-scale deployments, remote management is a key requirement and it contributes to system robustness as well. Approaches such as aggregate programming should be used to simplify the design, creation, and maintenance of complex IoT software systems. Here, the basic unit is no longer a single device, but instead a cooperating collection of devices: details of behavior and position and number of devices are largely abstracted away. This can be accomplished by a layered approach to programming complex services, building on foundational work on the composition of distributed systems and then on general mechanisms for robust and adaptive coordination.

Functionality and Processes

From a functional perspective, IoT systems in general are expected to collect sensor data, transport it to processing and decision-making locations, and enable the implementation of actionable insights. Real-time analytics will be a requirement involving business intelligence to machine learning, data mining, predictive analytics, condition monitoring, and visualizations. This often involves complex processes and possibly a model-driven approach would help to address the complexity. For example, business process modeling (BPM) is an established technique for modeling and executing complex processes in enterprises. The enterprise adoption of IoT technologies could be accelerated if these techniques are adapted to the requirements of real-world IoT.

Keep the End User in Focus

User experience is a key factor for the success of IoT applications. It is important to keep the user at the center to validate the IoT product’s value to the user. Mere technology “wow” factors cannot ensure the longevity of the device and application. The value proposition should consider social context and expectations as well. For example, driving behavior monitoring for a usage-based insurance premium had mixed acceptance with personal vehicles. At the same time, it had a taste of success in fleet-monitoring applications. Here for the same solution, the end user and the context were different. In the first case, the car owner is the end user who has the choice of enrolling in the scheme, and there is a shadow of privacy concern. For the second use case, the fleet manager is the end user and he or she has the value proposition and control to implement the solution.

Appropriate Business Model

The business model is a plan for the successful operation of a business, identifying sources of revenue, the intended customer base, products, and details of financing. A strong business model is the key to sustain any IoT application, and IoT has the potential to enable new business models. The World Economic Forum outlined four phases of adoption of industrial IoT applications in their report. They are operational efficiency, new products and services, outcome economy, and autonomous and pull economy. In the initial phase, focus can be given to improving operational efficiency through various means such as improved asset utilization, operational cost reduction and worker productivity. In the second phase, new products and services can be focused. This may involve new software services, data monetization, and payment models such as pay per use. Outcome economy will focus on systems by forming and leveraging new connected ecosystems, platform-enabled marketplaces, and new payment processes. Finally, autonomous and pull economy is leveraged by continuous demand sensing, end-to-end automation, resource optimization, and waste reduction.

In each of these phases, there is a potential to explore appropriate new business models. Remote monitoring and maintenance are an upcoming service and business model enabled by IoT. New payment models that will enable charging for usage and service quality levels are promising. “TotalCare” aerospace service from Rolls Royce is an example of a business model applicable to IoT where the payment mechanism for aircraft engines is $/engine flying hours. Supporting such a service requires extensive sensing, monitoring, analytics, and prediction.

Develop an Ecosystem

For an IoT application to succeed, there should be an ecosystem to support in terms of technology awareness, skill set, developer community, and third-party services. Platform-based offerings should consider providing easy access to third-party developers. Opensourcing is being increasingly adopted for some of the horizontal technologies those are hard to crack, and it may result in a robust solution with the contribution from the diversity of the community.

---

Click here to view the complete list of Methods & Tools articles

This article was published in December 2017