Data science and machine learning in wastewater intelligence 🌳
summary
Kando installs proprietary sensors and streams time series to near‑real‑time pipelines that flag pollution spikes, use semi‑supervised learning plus expert labels to expand scarce annotations, and apply regression models to replace costly ground‑truth devices — often achieving orders‑of‑magnitude cheaper coverage. Deployment optimizes manhole placement on a sewer graph, localization reverses signal diffusion and dilution with signature matching, and tools like tslearn support clustering and anomaly detection for actionable alerts.
AI-written summary · about
chapters (5)
- 0:00Introduction and welcome
- 0:26What is wastewater intelligence
- 8:34Data sources and machine learning
- 16:21Sensor deployment strategies
- 23:46Key takeaways and conclusion
related talks
18mData science and machine learning in wastewater intelligence
26mFrom Data to Decision: Monitoring and Improving Water Networks using AI/ML
08mUsing machine learning to infuse data operations
37mBig Data Analytics, Machine Learning, Real Time Processing
20mAutomating Data Science for the IOT
35mData science and Internet of Things on the edge
37mFrom Sensors to Insights at the Edge: Applying AI to IoT Data
24mLeak Detection System using Machine Learning Techniques