Forecasting Bike Movements with IBM

Key Facts

Status: Completed

Connectivity: Bluetooth Cellular

Forecasting Bike Movements through Intelligent Reasoning

By Bora Caglayan, Applied Researcher at the Innovation Exchange, IBM Ireland Lab

The Innovation Exchange (IIX) team as part of the IBM Ireland Lab, is building capabilities that will ensure IBM remains a market leader for horizon technologies into the future. As a demonstration of this capability, we have collaborated with Dublin City Council and the Dublin Smart Docklands initiative to showcase the power of Artificial Intelligence, Cloud and IoT technologies in the future of cities.

The human race has long philosophized on how seemingly independent societal and environmental events and activities are linked. The ‘butterfly effect’, ‘chain reactions’, ‘root cause analysis’, ‘causality’ and many other concepts have been the focus of intensive research and investigation for decades. Understanding accurately what factors impact the performance of any system, economy or otherwise requires a 360° view of the interrelated environment. The unprecedented increased volume of data that has been generated in urban spaces and the advances in computing technology have provided us with opportunities to realize this vision.

Leveraging this opportunity and State of the Art technologies, the IIX have created an Intelligent Reasoning Platform that connects unobtrusive and ubiquitous sensing technologies, advanced data management and advanced Artificial Intelligence (AI) models. The platform provides a framework for the development of AI solutions to address city related challenges including data-driven urban planning.

The IIX team have demonstrated the potential for Intelligent Reasoning to address major problems faced by cities including telecommunications planning, urban planning and security in major cities including Dublin, Paris and New York City. One of the key features of the intelligent reasoning platform is the ease of portability to different use cases that require data driven reasoning in a matter of hours.

Dublin has one of the most successful bike-sharing systems in Europe. The Smart Docklands Initiative provided the Innovation Exchange the opportunity to partner with Smart Docklands members to build a forecasting model for bike movements across Dublin. The model could help city planners in two ways in promoting sustainable commuting in Dublin. Forecasting the seasonal and abnormal bicycle movements are important to plan the supply of bicycles in different zones. In addition, the bicycle commuter flow patterns can be used to plan and maintain the cycling infrastructure in Dublin.

The Intelligent Reasoning Platform used data generated by the intelligent bicycle lights with embedded GPS sensors developed by See.Sense, another Smart Docklands Partner. This dataset contained the movements of 470 volunteer cyclists who participated in the See.Sense demo program for a period of two months. We fused this dataset with auxiliary data sources such as weather information and dynamic region functionalities such as eating, business and night life to forecast the bicycle flow in Dublin City Centre. Figure 1 below shows a heatmap from the Demo which illustrates the predicted movement for the bicycles for the next hour, with the red shading indicating areas with the expected highest concentration of activity. The forecast horizon (i.e. the future time frame that is used for prediction) can be calibrated based on the use case.

Figure 1 Dublin Bike Locations Demo. The heat map summaries the predicted movement for the bicycles for the next hour.

The Intelligent Reasoning Platform brings several innovations over the State of the Art. One key improvement is the ability to fuse multiple data streams with different dimensions for forecasting models. Another improvement is the provision of planning functionality on top of the forecasting model to guide the resource allocation process and anomaly handling. Lastly, through our service-oriented architecture a solution can be calibrated for a new use case in a matter of hours.

As an example of another use case, the IIX team used the Intelligent Reasoning Platform for network demand planning in Paris and Manhattan. In this case, we predicted the network demand using several data streams including weather data, dynamic functionality data and network throughput data for mobile phones. On top of that, we implemented the planning module that uses the predictions for optimizing the location of the portable base stations and small-cell base stations in the city.

The Innovation Exchange (IIX) is an innovation group which is part of the IBM Ireland Lab located at IBM’s Technology Campus in Dublin. The IIX has identified IBM’s areas of interest from an industry perspective and is actively building partner networks as well as technical capabilities that will help the IBM Ireland lab to become innovation leaders within IBM globally as part of a vibrant Irish ecosystem. The IIX was launched four years ago by the Ireland Lab as an open innovation initiative which intended to realize the potential of cross industry convergence to stimulate economic, scientific and societal opportunity. The purpose of the initiative is to accelerate impact by connecting the scale and enterprise level capabilities of IBM’s technical and market insights with the subject matter expertise, agility and scientific excellence of the Irish ecosystem, the European ecosystem and beyond.

Check out our video

for more information or reach us directly (bora.caglayan@ibm.com).

The Intelligent Reasoning Platform originated from IBM Results arising from COGNET, a project which has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 671625”.

Do you find this project interesting? If you have any ideas of projects which might benefit the Docklands, let us know today

Get in Touch