A Dynamic Pipeline for Spatio-Temporal Fire Risk Prediction

ABSTRACT

Recent high-profile fire incidents in cities around the world have highlighted gaps in fire risk reduction efforts, as cities grapple with fewer resources and more properties to safeguard. To address this resource gap, prior work has developed machine learning frameworks to predict fire risk and prioritize fire inspections. However, existing approaches were limited by not including time-varying data, never deploying in real-time, and only predicting risk for a small subset of commercial properties in their city. Here, we have developed a predictive risk framework for all 20,636 commercial properties in Pittsburgh, based on time-varying data from a variety of municipal agencies. We have deployed our fire risk model on Pittsburgh Bureau of Fire's (PBF), and we have developed preliminary risk models for residential property fire risk prediction. Our commercial risk model outperforms the prior state of the art with a kappa of 0.33 compared to their 0.17, and is able to be applied to nearly 4 times as many properties as the prior model. In the 5 weeks since our model was first deployed, 58% of our predicted high-risk properties had a fire incident of any kind, while 23% of the building fire incidents that occurred took place in our predicted high or medium risk properties. The risk scores from our commercial model are visualized on an interactive dashboard and map to assist the PBF with planning their fire risk reduction initiatives. This work is already helping to improve fire risk reduction in Pittsburgh and is beginning to be adopted by other cities.

References

1. Susan Athey. 2017. Beyond prediction: Using big data for policy problems. Science Vol. 355, 6324 (2017), 483--485.
2. John Murray Brown. 2017. Grenfell fire tragedy prompts inspection of 4,000 UK tower blocks. Financial Times. (jun. 2017). https://www.ft.com/content/a6d17890--526b-11e7-a1f2-db19572361bb
3. Alex Chojnacki, Chengyu Dai, Arya Farahi, Guangsha Shi, Jared Webb, Daniel T Zhang, Jacob Abernethy, and Eric Schwartz. 2017. A Data Science Approach to Understanding Residential Water Contamination in Flint. In Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 1407--1416.
4. Christopher E Church and Amanda J Fairchild. 2017. In Search of a Silver Bullet: Child Welfare's Embrace of Predictive Analytics. Juvenile and Family Court Journal Vol. 68, 1 (2017), 67--81.
5. Joseph Clare, Len Garis, Darryl Plecas, and Charles Jennings. 2012. Reduced frequency and severity of residential fires following delivery of fire prevention education by on-duty fire fighters: Cluster randomized controlled study. Journal of safety research Vol. 43, 2 (2012), 123--128.
6. Eddie Copeland. 2015. BIG DATA IN THE BIG APPLE. Capital City Foundation. (2015). http://capitalcityfoundation.london/wp-content/uploads/2015/06/Big-Data-in-the-Big-Apple.pdf
7. Stephanie Cuccaro-Alamin, Regan Foust, Rhema Vaithianathan, and Emily Putnam-Hornstein. 2017. Risk assessment and decision making in child protective services: Predictive risk modeling in context. Children and Youth Services Review (2017).
8. Marc DaCosta, Jeremy Krinsley, and Brian Abelson. 2015. Optimizing Local Smoke Alarm Inspections with Federal Data. Bloomberg Data for Good Exchange (2015).
9. Ben Green, Alejandra Caro, Matthew Conway, Robert Manduca, Tom Plagge, and Abby Miller. 2015. Mining administrative data to spur urban revitalization Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 1829--1838.
10. Katherine Hillenbrand. 2016. Predicting Fire Risk: From New Orleans to a Nationwide Tool. Data-Smart City Solutions (2016). https://datasmart.ash.harvard.edu/news/article/predicting-fire-risk-from-new-orleans-to-a-nationwide-tool-846
11. Jonathan Jay. 2017. How we predicted building fires in Baton Rouge, LA -- working version. Urban Analytics Blog (2017). https://scholar.harvard.edu/jonjay/blog/how-we-predicted-building-fires-baton-rouge-la-working-version
12. Karen Lavi, Ritvik Kharkar, Mathew Kiang, Christoph Hartmann, Paul Van Der Boor, Adolfo De Unanue, Leigh Tami, Anson Turley, Cedric Robinson, Brandon Crowley, Eric Potash, and Rayid Ghani. 2017. Using Machine Learning to Improve Emergency Medical Dispatch Decisions Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM.
13. Michael Madaio, Shang-Tse Chen, Oliver L Haimson, Wenwen Zhang, Xiang Cheng, Matthew Hinds-Aldrich, Duen Horng Chau, and Bistra Dilkina. 2016. Firebird: Predicting Fire Risk and Prioritizing Fire Inspections in Atlanta Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 185--194.
14. Michael Madaio, Oliver L Haimson, Wenwen Zhang, Xiang Cheng, Matthew Hinds-Aldrich, Bistra Dilkina, and Duen Horng Polo Chau. 2015. Identifying and Prioritizing Fire Inspections: A Case Study of Predicting Fire Risk in Atlanta. Bloomberg Data for Good Exchange, New York, NY, USA (2015).
15. Parisa Moshashaei, Seyed Shamseddin Alizadeh, et al. 2017. Fire Risk Assessment: A Systematic Review of the Methodology and Functional Areas. Iranian Journal of Health, Safety and Environment Vol. 4, 1 (2017), 654--669.
16. City of Pittsburgh. {n. d.}. Fire Prevention Code of Ordinances. (. {n. d.}). https://www.municode.com/library/pa/pittsburgh/codes/code_of_ordinances
17. Eric Potash, Joe Brew, Alexander Loewi, Subhabrata Majumdar, Andrew Reece, Joe Walsh, Eric Rozier, Emile Jorgenson, Raed Mansour, and Rayid Ghani. 2015. Predictive modeling for public health: Preventing childhood lead poisoning Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 2039--2047.
18. Mahsa Salehi, Laura Irina Rusu, Timothy Lynar, and Anna Phan. 2016. Dynamic and robust wildfire risk prediction system: an unsupervised approach Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM, 245--254.
19. Erika Salomon, Tzu-Yun Lin, Kate Boxer, Hareem Naveed, Lauren Haynes, Joe Walsh, Jen Helsby, Steve Yoder, Robert Sullivan, Chris Schneweis, et al. {n. d.}. Reducing Incarceration through Prioritized Interventions. University of Chicago Data Science for Social Good (. {n. d.}).
20. Chengcheng Xu, Andrew P Tarko, Wei Wang, and Pan Liu. 2013. Predicting crash likelihood and severity on freeways with real-time loop detector data. Accident Analysis & Prevention Vol. 57 (2013), 30--39.

Index Terms

• Computing methodologies

  • Machine learning

    • Machine learning approaches

      • Classification and regression trees