MAD - MAX | Devpost

Inspiration

Our passion for robotics and problem-solving inspired us to take on the MAD - MAX project at UTRA Hacks. Navigating with only a single colour sensor and an ultrasonic sensor posed a significant intellectual challenge, and we were eager to push the limits of what we could achieve.

What it does

MAD - MAX is a pathfinding robot that autonomously navigates its environment using only a color sensor, an ultrasonic sensor, and two DC motors. It was designed to complete three specific challenges set by UTRA Hacks, requiring precise movement, obstacle avoidance, and adaptive pathfinding.

How we built it

We built our system around an Arduino Uno R4 Minima, which controlled two DC motors through an L298N motor driver and was powered by a 9V battery. A colour sensor was used to detect path markers, while an ultrasonic sensor helped with obstacle avoidance. Additionally, we designed and integrated a 3D-printed claw to interact with objects as part of the challenges.

One of the biggest hurdles we encountered was the absence of an IMU or gyro for heading estimation. Without these, maintaining a consistent direction proved challenging, often leading to unpredictable navigation. To overcome this, we experimented with multiple pathfinding algorithms and ultimately adopted a collapsing distance algorithm, which progressively reduced the error between the robot’s position and its target.

For the second challenge, we had to translate colour sensor data into directional movements, allowing the robot to interpret and follow paths dynamically.

The third challenge required us to navigate without revisiting previously explored areas. Since we lacked onboard memory mapping or precise localization sensors, we implemented a time-based tracking approach to estimate previously visited squares. By monitoring the time elapsed during movements and interactions, we approximated our position relative to the start, ensuring efficient navigation while minimizing redundant paths—all without the need for advanced tracking hardware.

Challenges we ran into

Unreliable sensor readings: The colour and ultrasonic sensors were finicky and required frequent recalibration.
Lack of a heading sensor: Without an IMU or gyro, maintaining direction was difficult, requiring creative algorithmic solutions.
Hardware instability: We encountered frequent issues with power fluctuations and motor inconsistencies.