Design of an All-Purpose Terrace Farming Robot

Vibhakar Mohta, Adarsh Patnaik, Shivam Kumar Panda, Siva Vignesh Krishnan, Abhinav Gupta, Abhay Shukla, Gauri Wadhwa, Shrey Verma, Aditya Bandopadhyay

Abstract

Terrace farming is a critical agricultural practice for soil and water conservation in hilly terrains, yet it remains one of the most labor-intensive sectors due to the inability of traditional machinery to navigate steep, narrow steps. This paper presents the design and implementation of Aarohi, an autonomous modular robot engineered to bridge this automation gap. Aarohi utilizes a novel scissor-lift climbing mechanism, optimized via genetic algorithms, to traverse terrace steps up to 40 cm in height. Equipped with a modular "Ploughing, Sowing, and Rolling" (PSR) suite and an adaptive control stack, the robot demonstrates high stability and navigational precision on unstructured terrain.

Introduction

In mountainous regions, agriculture is constrained by the physical geometry of the land. Traditional tractors are too large for narrow terrace strips, and standard wheeled or tracked robots often fail when encountering the vertical "risers" between terrace levels. While aerial drones offer solutions for spraying, they lack the payload capacity and ground contact required for heavy-duty tasks like tilling or seeding.

Aarohi was developed to address these specific challenges. By focusing on vertical mobility and modularity, the platform provides a versatile solution for small-scale farmers, reducing the physical burden of manual labor while maintaining the structural integrity of the terrace environment.

Mechanical Design and Optimization

The Scissor-Lift Mechanism

The defining feature of Aarohi is its dual scissor-lift chassis. Unlike fixed-axle robots, Aarohi can independently adjust the height of its front and rear wheel assemblies. This allows the robot to "climb" vertical obstacles by lifting its leading wheels onto a step, followed by its trailing wheels, while maintaining a stable center of mass.

To ensure the mechanism could handle the high forces required during the initial lift phase, the researchers employed a Multi-Objective Genetic Algorithm (NSGA-II). This optimization process balanced two competing objectives:

  1. Minimizing Actuator Force: Reducing the peak load on the 1000N linear actuators.
  2. Maximizing Stroke Length: Ensuring the robot could reach a vertical height of 40 cm.

Structural integrity was verified through ANSYS simulations, confirming that the scissor links could support a 40 kg operational load without significant deformation.

Modular PSR Suite

The robot features a detachable "Ploughing, Sowing, and Rolling" (PSR) module. This integration allows for three distinct agricultural operations in a single pass:

  • Ploughing: A disc plough with optimized tilt and disc angles for hilly soil.
  • Sowing: A stepper-motor-controlled seed metering system for precise spacing.
  • Rolling: A trailing roller to firm the soil over the seeds, ensuring proper germination conditions.

Control Architecture and Navigation

Aarohi’s software stack is built on the Robot Operating System (ROS), utilizing a multi-sensor fusion approach for localization and stability.

Localization and Perception

An Extended Kalman Filter (EKF) fuses data from wheel encoders, an Inertial Measurement Unit (IMU), GPS, and 1D Lidars. The Lidars serve a dual purpose: they act as a "fail-safe" by detecting terrace edges to prevent falls and provide real-time width estimation of the terrace to facilitate autonomous U-turns.

Adaptive Control Strategies

Navigating loose soil on a slope introduces significant challenges in pitch stability and side-slip. To counter this, the team implemented two primary adaptive algorithms:

  1. Adaptive PID (Pitch Control): Based on the MIT rule, this controller adjusts the scissor-lift extension in real-time to compensate for shifting weights (e.g., as seed or water tanks empty) and uneven terrain.
  2. Adaptive Pure Pursuit: This navigation algorithm accounts for side-slip on inclines, allowing the robot to converge to a zero cross-track error and maintain straight sowing lines even on slippery or soft soil.

Experimental Results

The Aarohi prototype was tested in a simulated terrace environment, yielding the following performance metrics:

  • Vertical Mobility: Successfully climbed steps up to 40 cm, doubling the capability of standard wheeled platforms of similar scale.
  • Load Capacity: Maintained mechanical stability with payloads up to 50 kg.
  • Navigational Precision: The Adaptive Pure Pursuit controller successfully eliminated cross-track error, ensuring consistent row spacing for crops.
  • Operational Efficiency: By combining three tasks into the PSR module, the robot significantly reduced the energy and time requirements compared to sequential manual operations.

The project’s efficacy was further validated by winning First Position at the 8th Inter IIT Tech Meet’s Autonomous Terrace Farming event.

Discussion and Future Work

While Aarohi represents a significant step forward, several areas for development remain. The current Lidar-based perception system is movement-dependent; future iterations could incorporate stereo vision for more robust, static environment mapping. Additionally, while the robot’s traversal is autonomous, the high-level agricultural strategy—such as determining optimal sowing depth based on soil moisture—currently requires human input. Integrating on-board soil sensors could lead to a fully closed-loop autonomous farming system.

In conclusion, Aarohi demonstrates that through optimized mechanical design and adaptive control, it is possible to bring the benefits of modern robotics to the challenging and vital domain of terrace farming.

Summary updated

Abstract

Automation in farming processes is a growing field of research in both academia and industries. A considerable amount of work has been put into this field to develop systems robust enough for farming. Terrace farming, in particular, provides a varying set of challenges, including robust stair climbing methods and stable navigation in unstructured terrains. We propose the design of a novel autonomous terrace farming robot, Aarohi, that can effectively climb steep terraces of considerable heights and execute several farming operations. The design optimisation strategy for the overall mechanical structure is elucidated. Further, the embedded and software architecture along with fail-safe strategies are presented for a working prototype. Algorithms for autonomous traversal over the terrace steps using the scissor lift mechanism and performing various farming operations have also been discussed. The adaptability of the design to specific operational requirements and modular farm tools allow Aarohi to be customised for a wide variety of use cases.
BibTeX
@inproceedings{mohta2022design,
  author = {Vibhakar Mohta and Adarsh Patnaik and Shivam Kumar Panda and Siva Vignesh Krishnan and Abhinav Gupta and Abhay Shukla and Gauri Wadhwa and Shrey Verma and Aditya Bandopadhyay},
  title = {{Design of an All-Purpose Terrace Farming Robot}},
  booktitle = {MESA},
  year = {2022},
  url = {https://openreview.net/forum?id=0pkD0tI8Od},
}
Figure