Presented By O'Reilly and Cloudera
Make Data Work
Sept 29–Oct 1, 2015 • New York, NY

Oulu Smart City pilot

Susanna Pirttikangas (University of Oulu)
4:35pm–5:15pm Thursday, 10/01/2015
IoT & Real-time
Location: 3D 02/11 Level: Intermediate
Average rating: *****
(5.00, 2 ratings)

Oulu Smart City in Finland has a lively living lab tradition. We continuously collect data from transportation, infrastructure, and people; expand our ecosystem of data providing and consuming companies, research institutes, city officials, and citizens; and we develop services on top of the ecosystem.

To exploit our data lake, we need different technologies and algorithms to handle the streams of mobility-related data, aiming to reach capability for distributed real-time analysis and low latency ad hoc queries. In this talk, a run-through on real-world use cases will be presented, and for the use cases, we discuss the required components.

Our services include situational pictures for city dwellers, a real-time speeding alert system, interactive dashboards to visualize statistics about the traffic flow, a recommendation system for individual drivers, realization of distributed reasoning on traffic events, and our plans to experiment on the edges of the network with mobile code.

We discuss the pros and cons of different technologies we have tested in our use cases:

  • Lambda architecture with data collection and ingestion with Apache Kafka, and implementation of the batch layer using Hadoop File-system (HDFS), Apache Hive, and Apache Spark. The speed layer consists of Kafka combined with the Apache Storm stream processing engine. The serving layer in our platform consists of OpenTSDB and Apache Hbase.
  • Traffic data analytics platform implementing heterogeneous data aggregation from different sources: moving object data, weather, etc. to HDFS using Flume, and performing data filtering and real-time analysis with Spark Streaming alongside with Flume. The system periodically executes Spark (or MapReduce) jobs to analyze the data in HDFS.
  • Python and R for data analysis.
Photo of Susanna Pirttikangas

Susanna Pirttikangas

University of Oulu

Susanna Pirttikangas, D. Sc. (Tech.) received her PhD in embedded systems from the University of Oulu, Finland. Her post-doctoral visits were to Japan (Waseda University, 2004-2005 and Tokyo Denki University, 2008) and China (Tsinghua University, 2011). She is a co-leader of the Interactive Spaces research group within the Department of Computer Science and Engineering. The group is lead by Dean Jukka Riekki, and other co-leaders are Senior Research Fellow Mika Rautiainen and Iván Sánchez. In the team, Susanna works as a data scientist specializing in situation awareness. She has experience in developing methodology to de-noise, fuse, segment, and classify real-time data streams. Susanna has worked in different companies in research and development, and is an active member of the international research community as a workshop and conference organizer, as well as serving as a reviewer and PC member in top journals and conferences in her field.

Dr. Pirttikangas is a management group member of research in the Faculty of Information Technology and Electrical Engineering, University of Oulu, Finland. She has also has served as a governing board member of Infotech Oulu, an umbrella organization for information technology research at the University of Oulu. Currently, she works in the DIGILE D2I Strategic Centre for Science, Technology and Innovation (SHOK) Program (Traffic mini-ecosystem) and China-Finland ICT Alliance Everyday sensing project. At the University of Oulu, she lectures in the algorithms and data structures course and is actively participating in the development of the department’s curriculum (for example under the topic big data processing).