The rapid development in computing hardware, and sensor and user interface technologies enable the creation of more and more powerful computing solutions. Mobile, IoT, wearable and ubiquitous systems are designed to help us in our everyday life as well as in any profession. They heavily use machine learning components to automate and provide intelligent features. When working in concert with robot technology the solutions can also physically act in the world.
Software engineers developing such solutions are faced with a range of challenges to create useful solutions that consistently help users wherever they go, at an affordable cost and with a variety of hardware configurations. This means that successful solutions must deliver functionality that adapt to available resources at a given moment, e.g., in terms of sensor data accuracy, available energy, user interface options, processing power and communication options. To remain lawful, they also need protect privacy.
In this DIREC talk Mikkel Baun Kjærgaard will present and discuss engineering tactics to address such concerns and open-source software technologies that help implement tactics in practice. The tactics and software technologies are the result of more than 15 years of use-inspired basic research in projects with heavy industry involvement. The talk will give examples of ongoing projects researching tactics and software technologies for the design of systems with robotic elements. The talk will conclude with directions for future research.