Extending the Target-Search Algorithm for Kilobots by Adding Random Walk Behavior

This paper discusses the moving behavior of Kilobots, particularly target search and collision avoidance. When using minimalistic robots (e.g., Kilobots), the approach to develop an algorithm is limited by computational power and sensorial functionality. Therefore, Kilobots help examine the challenges of limited simple medical nanorobots. In this paper, an algorithm is presented which is an extension of target-search and collision avoidance by introducing random walk behavior using the Lévy distribution. The results show that the random walk behavior improves the movement of Kilobots in some respects. The robots tend to discover an unknown area more efficiently in a specific time. In addition, the robots are less likely to get stuck on boarders and obstacles. On the other hand, the introduced random walk does not improve the ability of target search and collision avoidance.