Abstract:

Programmable robotic sets have been used in many different educational contexts. From teaching fundamentals of robotics and control, programming, and informatics thinking through interdisciplinary scenarios including demonstration of phenomena in Physics and Mathematics, encouraging, and attracting girls to study science and technology, to different types of robot competitions which are very motivating events that foster learning, creativity and enhance the learners' exploration, cooperation, and communication skills. Throughout the decades, various programming paradigms have been considered by the main stream designers and software developers for the robotic sets: originating in the beautiful Seymour Papert's constructionist text programming language Logo, suitable for interactive and dynamic programmatic control of tethered models and unparalleled learning lessons on early experiences with data representation and manipulation, switching over to autonomous models with off-line program-download-run paradigm, starting with simple case tools with a support for simple global events, such as in Robotics Invention System for RCX units, followed by a long-term cooperation with National Instruments, the producer of the infamous LabVIEW software for instrumentation. This resulted in iconographic programming language with an unfortunately complex internal model resulting in many unstable versions of IDE in the early years of NXT control units, finally converging to all-unifying block-style programming akin the very popular MIT Scratch programming language that ruined the diversity of programming paradigms, while learners took advantage of an easier transition between various educational systems, however at the cost of losing the richness of the approaches. This rather epic development was nailed by wide spreading of the Python programming language use for programming educational robots, which due to its interpreted nature and dynamic memory allocation model tends to be completely unsuitable for programming low level and close-to-hardware electronics appliances that simple educational robot models undoubtedly are. We call for a renaissance in the diversity of programming styles and for exploration of a range of paradigms and approaches at all levels of the crucial low floor - high ceiling - wide walls spectrum of educational systems. In this work, we attempt to contribute to this effort in two independent ways: First, we propose, fully implement, demonstrate, and evaluate a new way of programming educational Spike Prime robots using enhanced model of finite state machines. Modelling robot behaviours using state automata proved to be a feasible approach and naturally suggests suitability of FSM programming model, which is superior to the Scratch-like scripting due to the state-local nature of events, and better overview of the robot interaction with its environment. Second, we propose and provide a set of educational projects of various difficulty for controlling robot with the Raspberry Pi Build Hat [1] as an alternate for the Spike Prime Hub expanding on the capabilities of this educational system with Internet connection, camera, powerful computational resource, possibility to connect Arduino and hundreds of low-cost sensors and actuators. The set of projects we developed pushes the ceiling of the capabilities of the Spike Prime system higher towards advanced projects, suitable for deployment in secondary technical schools and other advanced users’ milieus.

Keywords:

Robots, SPIKE Prime, programming language, Raspberry Pi Build HAT, projects.