Assessing the practicality of the robot platforms from Swarmfarm robotics
Abstract
This project assessed the SwarmFarm UGV for autonomous vineyard operations in Coonawarra, with a primary focus on inter-row slashing. Despite meeting core safety, navigation and security requirements, the UGVs performance was limited by overly sensitive obstacle detection, frequent stoppages and slow technical support, which restricted progress across several evaluation areas. Due to these issues, the trial ended early, preventing full testing of capabilities such as adverse-weather operation, autonomous decision-making and sensor integration. Financial modelling based on limited operational data showed a negative ROI for slashing alone, though future multifunctional use or multi-unit supervision may improve viability. Overall, the trial highlighted both the potential and current limitations of UGV technology for commercial vineyard use.
Summary
The SwarmFarm unmanned ground vehicle (UGV) was trialled in a vineyard in Coonawarra over a period of 14 months. Initially, the trial faced delays due to contracting issues, followed by setbacks related to COVID-19, which shifted the timeline from 2022-2024. The primary function of the UGV during this trial was inter-row vineyard slashing, and preliminary results suggest promising improvements in operational efficiency.
A comprehensive assessment was conducted, focusing on six key performance areas. The UGV successfully met three critical criteria: (1) Safety, (2) Navigation and Path Planning, and (3) Privacy and Security. However, components related to (4) Autonomous Decision Making and additional aspects of Navigation and Path Planning could not be fully evaluated due to time constraints. Specifically, the system was not tested on its ability to handle unexpected scenarios, such as fallen posts, pedestrians, animals, and other vehicles, nor was it assessed for its response to (5) Operate in Adverse Weather Conditions, including fog, rain, and nighttime operations.
The trial highlighted two major shortcomings of the UGV. First, the system lacked sensitivity in dynamic environments, where Autonomous Decision Making is essential. This limitation significantly impacted operational efficiency, requiring frequent interventions in areas with dense vegetation and canopies, which ultimately increased costs and disrupted other vineyard operations. Second, challenges related to access to (6) Maintenance and Support led to downtime, hindering the anticipated completion of the trial. Addressing these issues will be crucial for enhancing the UGV’s effectiveness in future applications.
Additionally, four other extra areas were assessed, though they were weighted less heavily due to the novelty and early-stage development of UGV’s in the wine industry. These areas were 7) Efficiency, 8) Sustainability, 9) Sensor integration, and 10) System intuitiveness. Nonetheless, these areas have the potential to significantly enhance productivity and efficiency, contributing to reduced operational costs while aligning with sustainability goals in the industry’s transition toward a lower carbon footprint. Expanding current UGV capabilities into these areas could lead to improved resource management, reduced labour requirements, and optimised workflows - key factors in today’s competitive agricultural landscape. Detailed insights into these considerations are provided in the document below, along with comments for clarity.
The project focused on analysing operational costs and estimating ROI, using existing vineyard machinery as a benchmark, while engaging industry partners through extension activities. However, the early trial termination limited data collection, preventing a full financial assessment. A modelled ROI based solely on slashing showed a negative return, indicating that using the UGV for slashing alone isn’t financially viable. This could improve if the UGV is used for other tasks, multiple rows are treated at once, functions are combined in a single pass, or one operator manages multiple units, all of which offer potential efficiency and labour savings.