1: \begin{abstract}

2: Distributed consensus has been intensively studied in recent years

3: as a means to mitigate state differences among dynamic nodes on a

4: graph. It has been successfully employed in various applications,

5: e.g., formation control of multi-robots, load balancing,  clock

6: synchronization. However, almost all existing applications cast an

7: impression of consensus as a simple process to iteratively reach

8: agreement, without any clue on possibility to generate advanced

9: complexity, say shortest path planning, which has been proved to be

10: NP-hard. Counter-intuitively, we show for the first time that the

11: complexity of shortest path planning can emerge from a perturbed

12: version of min-consensus protocol, which as a case study may shed

13: lights to researchers in the field of distributed control to

14: re-think the nature of complexity and the distance between control

15: and intelligence. Besides, we rigorously prove the convergence of

16: graph dynamics and its equivalence to shortest path solutions. An

17: illustrative simulation on a small scale graph is provided to show

18: the convergence of the biased min-consensus dynamics to shortest

19: path solution over the graph. To demonstrate the scalability to

20: large scale problems, a graph with 43826 nodes, which corresponds to

21: a map of a maze in 2D, is considered in the simulation study. Apart

22: from possible applications in robot path planning, the result is

23: further extended to robot complete coverage, showing its potential

24: in real practice such as cleaning robots.

25: \end{abstract}

26: