Robot Path Planning · Potential Field Method Simulator

The total potential at any point (X, Y) is the sum of attractive and repulsive potentials. The robot's planned path is the steepest descent down this combined field.

Where $U_{total}$ is the total potential, $U_{att}$ is the attractive potential from the goal, and $U_{rep, i}$ is the repulsive potential from the i-th obstacle.

The attractive potential is proportional to the square of the distance to the goal, creating a smooth, parabolic valley.

Here, $k_{att}$ is the attractive coefficient (set by the slider), and $d_{goal}$ is the Euclidean distance from the current point (x, y) to the goal. The force is the negative gradient: $\vec{F}_{att}= -k_{att}\cdot d_{goal}$, pointing directly toward the goal.

The repulsive potential is only active within a threshold distance $d_0$ from an obstacle, creating a steep, "divergent" hill.

$k_{rep}$ is the repulsive coefficient, $d_{obs}$ is the distance to the obstacle's edge, and $d_0$ is the influence radius. The force $\vec{F}_{rep}$ points directly away from the obstacle and becomes infinite as $d_{obs}$ approaches zero, ensuring collision avoidance.

Autonomous Mobile Robots (AMRs) in Warehouses: This method is foundational for real-time navigation of transport robots. The robot constantly recalculates the potential field from its current sensors, allowing it to dynamically avoid moving obstacles like forklifts or other robots while heading to a loading dock. The simulator's parameters directly correlate to tuning a real robot's "aggressiveness" ($k_{att}$) and "caution" ($k_{rep}$, $d_0$).

Spacecraft Debris Avoidance Maneuvers: For initial trajectory design, a potential field can model Earth (attractive for orbit) and debris clouds (repulsive). The influence radius $d_0$ defines a safe exclusion zone. While final trajectories use precise orbital mechanics, this method provides fast, intuitive "first-guess" paths, much like the path you see computed instantly in the simulator.

Robotic Surgery & Needle Steering: In minimally invasive procedures, a surgical tool must navigate around critical organs (repulsive obstacles) to reach a tumor (the goal). The potential field framework allows for pre-operative path planning and intra-operative adjustments, where organ positions might shift slightly.

Structural Topology Optimization Analogy: In CAE, designers use sensitivity fields to iteratively remove material from low-stress areas (repulsive) and reinforce high-stress areas (attractive) to find an optimal shape. The mathematical concept of following a gradient to an optimum is directly analogous to the robot's path planning process visualized here.

First, you might think that "increasing the coefficients will always improve performance," but this is a dangerous assumption. For example, setting the repulsive coefficient $k_{rep}$ extremely high makes the repulsive "hill" around obstacles too steep, which can cause robot vibration or prevent it from passing through narrow passages. Conversely, if the attractive coefficient $k_{att}$ is too strong, the robot may force its way right next to obstacles, increasing collision risk. Balance is key. You need situation-specific tuning , such as starting with values like $k_{att}=1.0, k_{rep}=0.5$ for wide warehouse aisles and reducing the repulsive influence radius $d_0$ for narrow passages.

Next, the misconception that "the calculations are light, so real-time updates are easy." While the computation per step is indeed light, if the robot gets trapped in a local minimum, it can fall into an "infinite loop" state, continuously calculating in the same spot. During implementation, safety measures like setting a maximum step count or adding small random perturbations when movement stops are essential.

Finally, this simulator's "point mass" model is a significant idealization. Real robots have size and shape. For instance, when planning for a 60cm-wide transport robot, you must consider obstacles as "inflated obstacles" by adding the robot's radius to the obstacle's radius. Trying this in the simulator by setting larger obstacles will help you intuitively grasp this concept.