Abstract
Problem Title: AI for workspace injury detection
Team Name: Ode to code
Team Members details:
Role Name USN College Email id Phone no.
(College Name
Reg no.)
Team Rakshitha M 1SI23CI041 SIT mrakshu@gma 9880679313
il.com
Leader
Team Anagha R 1SI23CI006 SIT 1si23ci006@sit. 8792695669
ac.in
Member 1 Mohan
Team Likitha V C 1SI23CI021 SIT 1si23ci021@sit. 9353711267
ac.in
Member 2
Team Ranjitha S R 1SI23CI045 SIT 1si23ci045@sit. 8904023585
ac.in
Member 3
PROBLEM STATEMENT:
This project aims to develop an AI-powered Workplace Injury Detection
System that leverages computer vision and sensor data to detect, analyze, and
alert stakeholders about potential workplace injuries in real-time. Workplace
injuries pose a significant risk to employee safety and productivity across
various industries, especially in manufacturing, construction, and warehousing.
Traditional safety monitoring systems rely on manual inspections and reactive
measures, often leading to delayed responses and increased injury rates.
Problem Description:
Workplace injuries are a major concern across industries such as manufacturing,
construction, logistics, and warehousing. Slips, trips, falls, improper lifting
techniques, and equipment-related accidents lead to significant financial losses,
reduced productivity, and, most importantly, harm to employees.
This project proposes the development of an AI-powered Workplace Injury
Detection System using computer vision and IoT-based sensors. Traditional
workplace safety monitoring methods rely on manual inspections, safety
compliance training, and post-incident reporting. These approaches often fail to
prevent accidents in real time due to human limitations and delays in response.
There is a growing need for an AI-driven solution that can continuously
monitor, detect, and alert supervisors about potential and actual workplace
injuries before they escalate.
Key issues to address:
Accurate Injury Detection – AI must correctly identify real
injuries and avoid false alarms.
Real-time Monitoring – The system should detect incidents
instantly and send quick alerts.
Privacy & Compliance – Must follow data protection laws and
respect employee privacy.
Integration with Existing Systems – Should work with current
CCTV, sensors, and alarms.
Handling Environmental Challenges – Works in different
lighting, crowded spaces, and various industries.
Scalability & Cost – Should be affordable and usable across
different workplaces.
Predictive Insights – AI should help prevent accidents before
they happen.
Employee Acceptance – Workers should trust and feel
comfortable with the system.
Proposed solution:
AI-powered Workplace Injury Detection System that uses computer vision and
IoT sensors to detect accidents like falls, slips, and unsafe behaviors in real
time. It sends instant alerts to safety personnel, enabling quick response, and
uses predictive analytics to prevent future incidents. The system ensures privacy
compliance, integrates with existing safety infrastructure, and provides data
driven insights to enhance workplace safety.
Key Features of the Solution:
1. AI-Powered Computer Vision – Detects falls, slips, and hazardous
movements from CCTV footage.
2. Real-Time Alerts – Sends instant notifications to supervisors via email,
app, or dashboard.
3. Cloud or On-Premise Deployment – Can be hosted on the cloud or
integrated with local servers.
4. Predictive Analytics – Uses historical video data to identify high-risk
areas and prevent incidents.
5. Privacy Protection – Implements face blurring and anonymization to
comply with privacy laws.
6. Dashboard & Reports – A web-based dashboard for tracking incidents,
safety trends, and analytics.
Use Case Diagram
(In this section, provide the use case scenario i.e. how your software is going to work. In this
section you should include use case diagrams as well. This would also include the work flow
of your software)
Technology Stack
AI & Computer Vision:
MediaPipe, OpenCV – Detect falls, slips, and unsafe movements.
Backend:
FastAPI (python) – Process video data and store detected incidents.
Frontend:
Streamlit– Display detections and alerts in a simple web interface.
Deployment:
Google Colab/Local Laptop – Run the AI model and process video streams.
Applications
1. Industrial Safety Monitoring – Detects falls, slips, and unsafe behaviors
in factories and warehouses.
2. Construction Site Safety – Monitors workers for accidents and ensures
compliance with safety protocols.
3. Healthcare & Elderly Care – Identifies falls and injuries in hospitals,
nursing homes, and care centers.
4. Retail & Office Spaces – Ensures employee safety in large workplaces,
reducing liability risks.
5. Public Safety & Surveillance – Detects accidents in public spaces like
train stations, malls, and airports.
6. Logistics & Transportation Hubs – Monitors loading docks and
warehouses for injury prevention.
7. Workplace Compliance & Risk Assessment – Helps companies analyze
safety risks and improve policies.
