������

There’s no shortage of hype when it comes to the Internet of Things. But many companies discover that getting projects up and running is hard work. In such instances, it’s worth remembering the Internet of Things is not the goal in itself: it’s a means of creating new business models and applications that will deliver high value for customers and employees alike.

As Antoine de Saint-Exupery wrote in his book : “If you want to build a ship, don't drum up men to collect wood and assign them tasks and work, but rather, teach them to long for the wide-open sea.”

When it comes to the implementation of data-driven business and service models for IoT, innovative approaches, cultural change, and a motivated team are needed, because IoT projects rarely fail due to technology or the budget. When they fail, it’s because of the existing corporate culture, lack of flexibility, silo structures, incorrect procedures and lack of know-how.

The greatest need today: Industry 4.0

But despite the challenges that companies face when it comes to IoT, many have ambitious plans for 2018/19. According to a study by analyst group IDG, IoT investment is primarily directed to projects in cloud services, security, and data security, followed by IoT hardware such as sensors, actuators or gateways.
According to the study, most IoT projects are currently looking at automation and data exchange in manufacturing technologies, followed by smart, connected products and predictive maintenance programs.

Software: the queen of the Internet of Things

Given the current trajectory, it’s not hard to see a future where software defines the performance of machines and factories. That will mean machine builders and software developers will need to cooperate closely — and that favors an agile approach. But this may be easier said than done: Today, in mechanical engineering, for example, we often see friction when modern agile software development encounters good hardware manufacturers, there are usually points of friction.

Software developers are faced with the challenge of using limited hardware resources. Meanwhile, machine builders need to understand that planned software cannot be pre-designed on paper. Here, both areas (Hardware and Software) need to move towards one another with the following learning objective: in the future, ‘digital and software’ are driving traditional business. This is because development cycles have shortened significantly, and faster deployments of smaller releases and apps are the current agenda. In the case of IoT projects, the gap between hardware and software development needs to be closed, and this is achieved with DevOps, among other things. You build it; you run it.

Take the example of household appliances such as washing machines or coffee makers: the innovation and added value now reside in the software and the intelligence. This trend is also evident in pulse watches; the hardware has changed over the years slightly, but by no means to the same extent as the software with its ongoing updates and ecosystems. Devices have been and are constantly getting smarter, and are being transformed from a pure watch to a fitness tracker, and from there, to a personal, networked health consultant.

Networking via smart ecosystems

As smart ecosystems evolve, consumer devices will have to ship with the right networking and automation functions, based on smart ecosystems. The basis for this is cloud computing, dynamic platforms, and intelligent algorithms. They connect networked devices, manage applications, process data, deliver reports and contain advanced analytics functionalities such as cluster analysis and machine learning.

Figure 1: An intelligent ecosystem

From a technical point of view, an intelligent ecosystem is a combination of an embedded system, a mobile system, and an information system.
The device is on the lower level, regardless of whether it is a coffee machine or a system in a factory. An embedded system implements the required functionality and logic in the device, turning it into an intelligent device.

The mobile system ensures connectivity and the data exchange, for example, with a mobile app as a component. The most common way to connect a device is to use the mobile network, local Wi-Fi, or use another device as a gateway by Bluetooth.

The information system contains the domain logic and acts as a link to external services and systems. It adds value by combining intelligent and connected devices with other services. These services can send additional data to the device to provide the user with a better user experience or to improve functionality. Conversely, the device can provide data for services in order to allow activities such as predictive maintenance.

Automated tests: basis for security and agile developments

The integration of products into complex ecosystems poses different challenges in terms of development, testing and operation. One prerequisite for overcoming the hurdles is automated software testing; an integral part of software development. “Agile methods and continuous delivery require automated testing,” says Michael Fait, Head of IoT at ������.
The main reasons:

1. Complex systems with many networked devices have too many interdependencies to be able to be tested manually
2. Agile methods — such as continuous delivery and frequent release cycles — require the continuous automated testing of every change to the software

This also applies to the exchange of data shared by ecosystem participants. Full security concepts are required here in order to avoid security gaps from arising in the communication system and to control remote access to the system. Connected devices collect different and often personal data and share it with other components in the system. At the same time, new attack vectors come into play. Corresponding security concepts must, therefore, be developed, expanded or completely rethought.

A smart ecosystem in practice

Fresh Energy’s business model is based on a smart ecosystem. The company describes itself as the “first digital electricity provider in Germany.” It has developed an intelligent and scalable ecosystem for its customers that consists of several individual components and services: these include power, hardware, analysis algorithms, mobile apps and customer service.

Figure 2: Fresh Energy's smart ecosystem

Fresh Energy customers now have full control and transparency about their power consumption. In addition: Based on the smart meter’s power consumption data, Fresh Energy can now use algorithms, machine learning, and pattern recognition to visualize not only a household’s total consumption but also the consumption of individual devices. And it can compare that data with other devices or households.

The integration occurs on Fresh Energy’s digital platform, which ������ helped to develop. The platform features fast and scalable web applications, and related mobile apps with state-of-the-art Java and JavaScript frameworks (React, Redux, D3, Spring, etc.). The infrastructure runs on Amazon Web Services (AWS).

The result: Fresh Energy's digital platform is able to quickly attract thousands of new customers and adapt them to existing and new processes and innovations. In addition, various business partners can be integrated.

Conclusion: agility, flexibility, and scalability are the focus of Smart Ecosystems. They represent the decisive added value for users, customers, and partners, by simplifying processes, cutting costs and enabling innovative services or significantly improving existing ones.

Figure 3: Put agile at the heart of smart ecosystems

At the same time, companies should ask themselves the question as to whether their products add value. One sure-fire method of identifying whether you’re on the right track is to actively involve employees and customers in product development. Their reactions will give you an early indication if you need to correct anything — or even cancel a project.

Agility and iterative development based on regular feedback and close observation are always worthwhile, especially as wrong decisions, misconceptions, undesirable developments or changes can always occur, but they are quickly discovered with agile methods and can be corrected without losing much time, money and reputation. The problem really isn’t that mistakes happen, but that they are not detected and corrected quickly enough.

Smart ecosystems: what it still comes down to:

• Avoid big budgets and long project durations
  Develop many small budgets, use many iterations in your projects, and shorten your release cycles significantly.
• Be distinct
  You cannot be better than your competitors if you use the same software, processes and products/services. Individuality and creativity enable innovation and leadership.
• Break up silos
  Strike the right balance between re-engineering legacy systems and adding new infrastructures. Allow IoT integration.
• Manage innovations
  Accept, foster and shape technical, methodological, cultural and other changes.
• Never be sure
  Discard the idea of knowing what your customers, employees or partners want or what the right technology is. Changes happen too quickly; always find out again when you want to make or initiate decisions.
• Hardware is important, but data and software are even more important
  It’s no longer the hardware that decides, but its data and the intelligent software behind it. The winner is the one who combines hardware and software to create new products and services.
• Develop your corporate culture
  Change and progress begin in the mind and are made possible by motivation, identification, participation, freedom of design, ambition, desire, and recognition. Agile, design thinking, continuous delivery, microservices, DevOps, etc.
• Re-professionalize your teams
  Find, develop and retain new talents that enrich your existing teams. Say goodbye to looking for what fits your “previous formula.”
• Track your business goals and act in a value-driven manner
  The most important business goal worldwide is increasing sales and profit, and retaining existing customers, winning new customers. And just keep developing what helps you reach your actual goals.

In summary: to be successful with IoT, the limiting and risky factor is not the Internet of Things itself, but the approach.