Many energy, manufacturing and chemical workshops provide essential product and services to the society to fight Covid-19, and their jobs are high risk and labour intensive. To provide a safe environment in these workshops, we propose a fast-deployable, automatic and intelligent scenario-based Personal Protective Equipment (PPE) detection system. The solution leverages camera video analysis, artificial intelligence (AI) models and smart edge / internet-of-things (IoT) device, to support real-time worker’s PPE detection for the workplace.
TECHNOLOGY FEATURES & SPECIFICATIONS
Technology Features and Specifications:
1. The AI models developed can
be retrained to suit any PPE types such as goggles, gloves etc.; and
support real-time scenario-based PPE detection
2. Edge computing technology to support fast PPE detection on the edge.
3. Non-compliance alerts.
4. Statistics reporting of non-compliance to facilitate investigation.
This AI-based vision inspection solution can be applied across most industry. The common applications could include:
detection of non-conformance in a high-risk / controlled environment;
surveillance of irregular patterns of personnel / objects in certain settings e.g. high security environment; or
generation of behavioural insights for further analysis and training purposes.
In the Fourth Industrial Revolution (Industry 4.0), manufacturing companies face new opportunities and big challenges as well. With increasing complexity of working environment and rising cost of workers, workplace safety issue has been heightened to a level higher than ever, and companies are struggling to reduce worker injuries, working hours loss, compensation payment and strict offence penalty.
In Singapore, government targets to reduce fatal injuries to below 1.0 per 100,000 workers by 2028, a safety level that has been achieved by only four OECD countries on a sustained basis.
This solution can support energy, manufacturing and chemical workshops to leverage innovative technology to create a safe workplace solution using a smart machine-learning approach, which is based on vision AI and IoT edge computing technology.
The system can be widely applied in many essential services and high-risk plants in Singapore. It has three benefits:
1. PPE detection is conducted automatically without human contact, so it is safe for system operation.
2. PPE AI model can be retrained to other scenarios.
3. Scenario-based PPE detection expands the application of the solution to avoid false detection, and thus provide versatility of proposed solution to support different use cases.