Data Related Challenges to Deploying ML Systems in Production: An Example from Sewer Inspection

Biswas, R., Nebel, V. & Werth, D.(2024)
In: Arai, K. (eds) Advances in Information and Communication. FICC 2024. Lecture Notes in Networks and Systems, vol 919, S. 218-225, DOI:

Link zur Publikation:


Machine learning is one of the key techniques powering the digitization and AI transformation of our society. Steady interest, consistent effort and innovation from academia has transformed the field into one of the primary factors leading the fourth industrial revolution, Industry 4.0. A wide variety of businesses are now using machine learning for gaining insights from their workflows and solving real world problems. However, successful deployment of machine learning solutions in production is filled with challenges spanning a range of different issues. In this paper, we focus on the different data challenges encountered in deploying a machine learning system in the real world. We also discuss the data related issues we face in deploying an AI system that performs automatic defect identification on real world sewer videos and propose ways of handling such challenges.

Data Related Challenges to Deploying ML Systems in Production: An Example from Sewer Inspection | SpringerLink