Introduction

Efficiency at all levels of the Manufacturing Process

Simply put, Manufacturing is a simple game – the game of time vs. output – otherwise known to us as efficiency while achieving high level quality goals. Tech Mahindra’s AI-first approach helps manufacturers leap ahead in the game without having to sacrifice the human-first approach, rather helping the team focus on better and far more intelligent tasks.

• A lot of man-hours and effort are wasted in doing repetitive jobs, that can be utilised in more productive work
• In Manufacturing, mechanical accuracy is paramount – both for efficiency and for safety – this can be best addressed through AI
• Manufacturing is responsible for generation of huge amounts of data - to be able to quickly assemble, structure and process this data becomes quite difficult

• Assessment and planning to help set up new manufacturing centres of excellence
• Deep Data Analysis to establish manufacturing demand forecast
• Computer AI Vision to identify, read and analyse machinery and report their status
• Advanced analytics (Machine Learning) to obtain, structure, read and present manufacturing data accurately and efficiently

• For a reputed aircraft carrier, our AI-First solution helps in identifying potential issues with airflow over the wing at design time
• Our AI-First solution DECITM helps manufacturers identify wear and tear in machinery and suggest plausible measures
• Our cognitive ML solution can suggest the most viable concessions to engineers in the building process

Solutions

Aircraft testing is a vital matter, thus there is no scope of error. Our intelligent computer vision solution for gauging airflow analyses data on deviation in air cones behaviour patterns during aircraft testing and gives the testers detailed and insightful help in decision-making.

While inspecting manufacturing machinery for wear and tear, drones can reach places and show up-close angles, which might be manually impossible and dangerous too for humans to explore. Our Yolo model trained AI-enabled drones are capable of thorough inspection and providing live feed, which can help in identifying damaged parts efficiently.

Improving upon the existing ASIS process, we used deep knowledge discovery techniques to learn concessions-related parameters from historic data so that our ML model can suggest the most viable solutions to engineers.

Benefits

• Improved accuracy of test data
• Multifold decrease in manual efforts and efficient resource utilisation

• Improved mechanism for maintenance of damaged parts
• Accurate insights from live feeds
• Reduced human intervention

• Better turnaround time for concessions
• Opex reduction – up to 40%
• Increased volume of concessions managed at shorter time
• Reduced engineer’s intervention and man-hours