Enabling new business models for the era of physical digital convergence

Technology does not add value per se. Rather product developers need to understand the situation of their customers. With methodologies from human centered design, like empathy maps, this understanding can be developed and new product designs can specifically address the pains and gains of the customers.

Understanding humans is however not enough, also the business model needs to be scrutinized. Are customers still willing to invest in the product or is the product just a means to an end, which can be more attractively priced as a service? A competitive advantage often results from the combination of new technology with a new business model.

To combine traditional design and development of electro-mechanical products with software requires new approaches to system architecture, where the traditional product development process needs to be combined with the agile software development process and aligned with Systems Engineering practices.

The design and development of smart products requires agility beyond agile software development. Especially when new business models are tested or new customer segments addressed, a fast feedback loop needs to be established so that issues with the product-market are identified and corrected as soon as possible. At the same time, the often diverging lifecycles of software and hardware need to be synchronised.

The more connectivity a product provides, the more attack vectors for hackers exist. Hence, smart products need to be designed with defense-in-depth in mind. Anomalies in the operation of the product need to be identified and over-the-air software updates need to be used to patch vulnerabilities as soon as they have been identified.

With more data becoming available from sensors and user interactions, the functionality of smart products can be increasingly implemented using artificial intelligence. For this, data engineering and data science capabilities need to be developed in the respective product domains and a proper lifecycle management for training, testing, deploying, and operating the AI models need to be established.