A Semi-Automated Systematic Literature Review (SLR) Automation Tool for AI & Plant Imaging Research
Plant Stress AI Review Bot is a Python-based automation framework designed to perform large-scale systematic literature searches using programmable GUI bots.
It was originally developed to support the publication:
Walsh, J.J., Mangina, E., & Negrão, S. (2024). Advancements in Imaging Sensors and AI for Plant Stress Detection: A Systematic Literature Review. Plant Phenomics. https://doi.org/10.34133/plantphenomics.0153
The Plant Stress AI Review Bot automates repetitive database search tasks across multiple journal platforms using coordinate-driven interaction via PyAutoGUI.
The Plant Stress AI Review Bot enables researchers to:
- 🔍 Perform automated keyword-based database searches
- 📚 Systematically query multiple journal repositories
- 📊 Log study counts into structured spreadsheets
- 🗂 Archive search outputs into Zotero
- 🔁 Iterate thousands of search combinations
- 🧠 Reduce human error in repetitive SLR workflows
This bot was designed specifically for systematic review automation — not general web scraping.
Instead of manually performing thousands of search queries across academic databases, the Plant Stress AI Review Bot:
-
Uses predefined keyword groups:
- Root keys (e.g., abiotic stress)
- Parent keys (e.g., hyperspectral imaging)
- Child keys (e.g., supervised learning)
-
Iteratively constructs search strings.
-
Navigates database interfaces via pixel-coordinate automation.
-
Records outputs into structured datasets.
The original SLR executed:
- 6 root keys
- 56 parent keys
- 14 child keys
- Across 4 databases
- Totaling 28,224 automated searches
From Bot.py:
- Automation Engine: PyAutoGUI
- Clipboard Handling: Pyperclip
- Timing Control: time
- Execution Environment: Ubuntu (tested), adaptable to any OS
This bot is resolution-dependent and must be customized to your screen.
git clone https://github.com/YOUR-USERNAME/plant-stress-ai-review-bot.git
cd plant-stress-ai-review-botpython install.pyOr manually:
pip install pyautogui pyperclipThis bot will not work out-of-the-box without customization.
Before running:
Inside Bot.py, adjust:
- Screen resolution values
- X/Y click positions
- Row/column offsets
- Scroll positions
- Button locations
These must match your monitor and database layout.
The bot was originally built using:
1920 x 1080 resolution
Ubuntu OS
OneSearch database
If using another database (e.g., Google Scholar), you must redesign the navigation logic.
Inside the script:
bot.PAUSE = 1.0
bot.FAILSAFE = False- Increase
PAUSEif pages load slowly. - Enable
FAILSAFE = Truefor safety (move mouse to top-left corner to abort).
Once configured:
python Bot.pyThe bot will:
- Navigate to target applications
- Perform database searches
- Extract result counts
- Log outputs into spreadsheets
- Continue iteratively
Plant-Stress-AI-Review-Bot/
│
├── Bot.py
├── install.py
├── README.md
└── LICENSE
- Fully dependent on screen resolution
- Sensitive to UI layout changes
- Not robust against CAPTCHA systems
- Requires Zotero pre-configured (if archiving enabled)
- Designed for academic database interfaces
This is not an API-based scraper — it is GUI automation.
This tool interacts with live journal databases.
Users are fully responsible for:
- Complying with database terms of service
- Ensuring institutional access permissions
- Avoiding excessive query rates
The authors accept no liability for misuse.
If you use this bot in academic research, please cite:
Walsh, J.J., Mangina, E., & Negrão, S. (2024). Advancements in Imaging Sensors and AI for Plant Stress Detection: A Systematic Literature Review. Plant Phenomics. https://doi.org/10.34133/plantphenomics.0153
Contributions welcome.
- Fork the repository
- Create a new branch
- Submit a pull request
University College Dublin School of Biology & Environmental Science and the School of Computer Science