Automatic Training of Rat Cyborgs for Navigation

Abstract

A rat cyborg system refers to a biological rat implanted with microelectrodes in its brain, via which the outer electrical stimuli can be delivered into the brain in vivo to control its behaviors. Rat cyborgs have various applications in emergency, such as search and rescue in disasters. Prior to a rat cyborg becoming controllable, a lot of effort is required to train it to adapt to the electrical stimuli. In this paper, we build a vision-based automatic training system for rat cyborgs to replace the time-consuming manual training procedure. A hierarchical framework is proposed to facilitate the colearning between rats and machines. In the framework, the behavioral states of a rat cyborg are visually sensed by a camera, a parameterized state machine is employed to model the training action transitions triggered by rat's behavioral states, and an adaptive adjustment policy is developed to adaptively adjust the stimulation intensity. The experimental results of three rat cyborgs prove the effectiveness of our system. To the best of our knowledge, this study is the first to tackle automatic training of animal cyborgs.

Figure 1

Rat cyborg.

Figure 2

Manual training procedure.

Figure 3

Automatic training procedure.

Figure 4

Automatic rat cyborg training framework.

Figure 5

Motion parameters calculated by our method. The red point is the body position (P_b), the yellow point is the head position (P_h), and the blue line is the heading direction of the rat cyborg (Θ_r).

Figure 6

(a) An overhead view of the modified eight-arm radial maze and state division. A: at the tail end of the maze arms. B: in the passage of the arm and going inside. C: in the passage of the arm and going outside. C_L: in the center circle and the current task is to turn left. C_R: in the center circle and the current task is to turn right. D: in the center circle of the maze. E: out of the maze. (b) Task model. WA: other 5 wrong arms, f: FORWARD, l: LEFT, and r: RIGHT.

Figure 7

Automatic rat cyborg training system.

Figure 8

A side view of the modified eight-arm radial maze for automatic training. Two high plexiglass walls (40 cm × 15 cm) have been added to each arm (60 cm × 15 cm) to prevent the rat cyborg from climbing up the arm, and the width of each arm is set small enough to prevent the rat cyborg from turning around. Because the plexiglass walls is shorter than each arm, the rat cyborg can turn around at the tail end.

Figure 9

Navigation test maze. This maze is made of wood and comprises 10 × 10 unit squares (15 cm × 15 cm per unit square). The walls of the maze are 15 cm high and the outside walls enclose the entire maze. The starting point and four goals are also presented.

Figure 10

Learning curves of the automatic training.

Figure 11

Behavior changes in the automatic training.