By Mahesh Subramanian
Artificial intelligence has started to play a pivotal role in quality control across a variety of industries, transforming how inspections, maintenance, logistics, etc. are all done. What AI enables is predictive, objective and safe quality measures that frees up human time for higher order tasks and also helps speed up the whole process.
Humans inherently have subjective bias, and this is especially true in a manual quality setup wherein the quality parameters will change from shift to shift depending on the humans manning that particular station. Studies have also shown that the mistakes in quality control increases towards end of shifts as fatigue sets in. This is where automated systems with AI can provide an even, objective, fast and highly scalable solution.
AI systems improve with time unlike humans when presented with repetitive tasks. Most of the quality control comes under the repetitive category. Automated AI systems can and will be much faster than humans and also have a higher order of accuracy to detect even the minutest of defects under controlled conditions. With Industry 4.0 and also manufacturing moving to Just In Time setups, it becomes imperative for QC processes to enable proper quality checks across the whole supply chain and also manage the whole track and trace for inspections until the final product is in the hands of the end customer.
AI can not only find the defects and damages but can also do root cause analysis so that the cause can be identified upstream and fixed which will help reduce costly returns once the product leaves the manufacturing facility. To automate process that require human level intelligence we need systems and technologies that have this capability. There has been tremendous improvement in automation and AI technologies that can tackle myriad quality processes.
One such technology area is the use of computer vision for AI. With a good AI vision product dedicated to quality control, and training images that depict good parts and not good parts, training an AI can be quite fast and can be deployed real quick. In product environments that keep changing, use of AI is paramount so that multiple products can be tested without too many delays in the process. The advances in Convolutional Neural Networks (CNN’s) have enabled faster training of advanced image sets for faster deployment. Also technologies such as synthetic data generation as well as transfer learning from adjacent domains ensure that the system learning is deep, fast and highly scalable.
Given the current pandemic situation there is a pressing need for automation and AI enablement of quality processes. Automation with computer vision eliminates the need to have people in close quarters in a quality inspection station thus ensuring that safety is taken care of. It also ensures that there is no change to the existing workflow and space requirements as most of these systems can be retrofitted into existing quality bays. This helps in speeding up the digitization process whilst at the same time having deep actionable analytics that can help improve the whole process.
So to recap:
The benefits of automated visual inspection are:
- Enabling early error detection in the manufacturing processes and helping to ensure quality of the item before it’s moved to the next step
- Helping to gather historical and production statistics that can used to improve manufacturing capabilities
- Help reduce material waste, repair and rework costs, as well as added manufacturing labour expenses
Quality or lack of it usually has consequences. These are material consequences (rejections, replacements, etc.), financial consequences (warranty, late penalties, loss of margins and markets, etc.) and intangible consequences (customer dissatisfaction, loss of trust). For industrial companies, reducing these consequences is therefore a key issue for reducing costs and increasing performance. Companies that embrace AI and automation will reap both short and long term benefits.
(The author is Co-Founder & CEO at CamCom.ai)
