"Towards Value-Centric Big Data" e-SIDES Workshop - "Responsible Research: Analytics when dealing with personal and personalised mobility data: Track&Know", Prof. Ansar Yasar

"Towards Value-Centric Big Data" e-SIDES Workshop - "Responsible Research: Analytics when dealing with personal and personalised mobility data: Track&Know", Prof. Ansar Yasar

• 1.
  Big Data forMobility Tracking Knowledge Extraction in Urban Areas (Track&Know) Prof. Ansar Yasar University of Hasselt, BE 1 This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Grant Agreement No 780754. Towards Value-Centric Big Data: Connect People, Processes and Technology
• 2.
  Consortium 2Towards Value-Centric BigData: Connect People, Processes and Technology, Belgium, Brussels 2/4/19
• 3.
  3 T&K Vision Track&Know willresearch, develop and exploit a new software framework that aims at increasing the efficiency of Big Data applications in the transport, mobility, motor insurance and health sectors. Stemming from industrial cases, Track&Know will develop user friendly toolboxes that will be readily applicable in the addressed markets. T&K vision is in accordance with … • the EU Big Data Value Reference model • the US Big Data Reference Architecture model Source: NIST Big Data Interoperability Framework: Volume 6, Reference Architecture, v.2, Jun. 2018. https://doi.org/10.6028/NIST.SP.1500-6r1 Source: European Big Data Value Strategic Research Innovation Agenda (SRIA), v.4.0, Oct. 2017. http://bdva.eu/sites/default/files/BDVA_SRIA_v4_Ed1.1.pdf Towards Value-Centric Big Data: Connect People, Processes and Technology, Belgium, Brussels 2/4/19
• 4.
  4 • Platform tohandle heterogeneous streaming and archival data through the use of newly created toolboxes • Software Toolboxes, i.e. • data-intensive computing • interoperability • usability • standardisation • All data from streaming and archival data sources, as well as results from toolboxes can take full benefit of the computations of others, also taking advantage of seamless interoperability between their results. • Platform Architecture T&K Big Data Platform Towards Value-Centric Big Data: Connect People, Processes and Technology, Belgium, Brussels 2/4/19
• 5.
  5 • GPS locationdata from vehicle black boxes (provided by Octo Telematics) • Business cases: • Insurance: using historic telematics, environmental, demographic and geographic information, … gain in-depth and accurate crash probability estimation • Electric Cars: (i) cost-benefit of a switching to an electric car mobility; (ii) matching global charging times and charging points to drivers’ habits • Car Pooling: (i) park decreasing due to sharable routes; (ii) cost- benefit of switching to a sharing mobility paradigm; (iii) likelihood of finding a proper sharable route that matches time and geographical zone • Case studies: London (metropolitan city), Rome (metropolitan city), Tuscany, Italy (country-urban mixed area) Insurance Domain Pilot Sistematica S.p.A BIG DATA VALUE FORUM 2018 – Workshop 2.3 Policy issues, opportunities and barriers in big data-driven transport Vienna Austria 14/11/18 Towards Value-Centric Big Data: Connect People, Processes and Technology, Belgium, Brussels 2/4/19
• 6.
  6 • GPS andother sensor data (incl. fuel level and driver behavior data) from vehicles (tracks) and their drivers • Business objectives: • Predictive maintenance • Anomaly detection, reduction of false alarms • Correlation of Fleet Data with external Weather and Traffic services • Fleet costs reduction • Fleet downtime reduction • Fleet response time improvement • Improve driver behavior and reduce accidents • Datasets are geographically situated in Greece, Albania, Cyprus and few in various EU countries Fleet Management Domain Vodaphone Innovus Images produced by FRHF (Apr. 2018) Towards Value-Centric Big Data: Connect People, Processes and Technology, Belgium, Brussels 2/4/19
• 7.
  7 • A completeset of 5 years data from Obstructive Sleep Apnoea (OSA) patients • Business objectives: • Improvement of response times (increasing new patients, follow- up patients) • Reduction of unnecessary travel (reduce patient travel distances, courier costs, CO2 emissions • Cost efficiency gains • New methods of OSA diagnoses based on driver behaviour • Datasets are geographically situated in London area and Cambridge, UK • OSA increases risks of strokes and other cardiovascular diseases. • An OSA diagnoses results in the suspension of ones driving license Medical Service Optimisation (a) Network of Outreach Clinics staffed by PAP (b) Patient Journeys in 2018 to get access to OSA service Royal Papworth Hospital Towards Value-Centric Big Data: Connect People, Processes and Technology, Belgium, Brussels 2/4/19 (a) (b) Images produced by ILS (Jan. 2019)
• 8.
  Medical Pilot DataCollection • This pilots utilizes patient mobility information for two purposes: • Reallocating resources to better cover the catchment area based on geo- spatial analysis of pervious patient journeys and population propensity • Provide a proof-of-concept approach to (a) study the effects and (b) better diagnose obstructive sleep apnoea by collecting real-time driver behaviour using an App on the patients phone • Personal Information is not unavoidable it is required to get a clear understanding • Pseudo anonymisation is possible however enough points of information exist to make de-anonymisation trivial. 8Towards Value-Centric Big Data: Connect People, Processes and Technology, Belgium, Brussels 2/4/19
• 9.
  Measures in placeto ensure protection of Citizen Data • Better than industry standard security and encryption procedures • Very limited access to the data • High-levels of encryption of data at transmit, process and storage phases • Testing data for biases before making available on the platform • Ethical and informed processes to identify biases created by historic, socio- demographics, and geo-political factors • Informed consent of patients through: • Training patient facing staff to adequately inform patients of the aims and expected outcomes of the research, what data is being collected, all actors involved, and how the data is handled and processed • Involvement of citizen advocacy and support groups as active stakeholders in the research process to help identify the potential of disenfranchisement of citizens by adopted research practices • A robust and continuous ethical process review process to monitor data, tasks, outcomes and deliverables 9Towards Value-Centric Big Data: Connect People, Processes and Technology, Belgium, Brussels 2/4/19
• 10.
  Further Ethical Considerations •Un-identified significant data biases not detected or fully understood at this stage, but adversely affect results of our tools once ‘out in the wild’ – Significant commercialisation and exploitation plan • Duty of care to report occurrences of OSA behaviour found in non- medical pilot driver behaviour data to relevant authority VS affecting a professional drivers livelihood/affecting insurance rate of citizens. 10 WE DON’T HAVE ALL THE ANSWERS… YET Towards Value-Centric Big Data: Connect People, Processes and Technology, Belgium, Brussels 2/4/19
• 11.
  Thank You Project Managerand Point of Contact Dr Ibad Kureshi [email protected] Inlecom BVBA, BE 11 This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Grant Agreement No 780754.