Spearheading Video Analytics

Spearheading Video Analytics

Team members

Johan Farid Tang Wei Min (ESD), Wai Xin Ling Phyllis (ESD), Dong Yizhi (ESD), Yu Wenling (ESD), Joey Richie L Tan (ISTD), Prashanth Nair (ISTD), Zhou Yutong (ISTD)

Instructors:

Kenny Choo, Keegan Kang

Writing Instructors:

Grace Kong

Teaching Assistant:

Cheong Rui Zhi Jeremy, Anirudh Gajendra Rathi

Leveraging on the power of Artificial Intelligence, our project looks to re-innovate the process of quality assurance by involving video analytics in the identification of defects in products.

One key product provided by Schenker under contract logistics is a spectrum of value-added solutions to maximise customer satisfaction, to achieve above-market returns and to become the employer of choice. DB Schenker currently provides value-added services for customers through quality checks, packaging services, as well as labeling and price tagging.

For our project, we will be focusing on improving the quality inspection service by leveraging the power of Artificial Intelligence.

Through observations made during warehouse visits and interviews conducted with the inspectors, we identified 3 key areas of improvement to the current quality inspection process.

I. Physical strain on inspectors
The current quality inspection process is highly manual, time-consuming, and physically demanding of inspectors. Based on initial rounds of interviews with the inspectors, we found that up to 60% of the total time taken for the average inspection is being spent on the manual inspection of items as well as the documentation of inspection findings.

II. Varying objectivity in determining defects
In addition, human inspections are prone to varying acceptance standards, hence lack objectivity in defect detection.
III. Time-lag in communication between inspectors and manager:
Furthermore, it was identified that there are potentially large pools of time-lag accumulated in the waiting time whenever uncertainty in inspection arises, which would require intervention and final decisions to be made by quality assurance managers. This issue is further augmented with the Work-From-Home (WFH) arrangements currently put in place due to the COVID-19 situation.

Developed using React Native, chosen due to the stability of the platform.

Simple interface with no overcrowding of information on each page.

The software application serves 3 key purposes:

1) It guides the user into utilising the new workflow by adopting steps which are already familiar to the inspectors

2) It smoothen out the liaising process between inspectors and manager

3) It provides greater objectivity in inspection outcomes that still allows intervention from inspectors or manager if the need arises.

The realisation of the three functions is demonstrated through our video showcase.