The text on this page was automatically translated and hence may differ from the original. No rights can be derived from this translation.
That blockchain, IoT and data science are interesting technologies is no news. The fact that they can also be applied in healthcare and mobility was demonstrated by 19 students last Monday during the presentations of the results of Q1-2 2018 at the ICT Lab Utrecht.
Vaccination registration with Blockchain
The vaccination booklet, where travellers' vaccinations are registered, is in desperate need of an update. There is no coordination over the data currently being recorded. This leads to a fragmented approach, incorrect registration, and facilitates fraud. Also, losing your booklet can pose a significant problem (and it happens quite often...).
By digitising the vaccination booklet, the composition of the recorded data can be controlled. Additionally, digitalisation offers possibilities to register and authorise users of the booklet. Blockchain provides added value by eliminating the need to trust a central party to safeguard and modify the data.
Ana, Joris, Karim, Lex and Timo explored possibilities for digitalising the vaccination booklet using blockchain. The results of their demo, presented live on Monday, were so promising that they will dedicate themselves to building a functioning and tested application from September onwards.
Big Data in healthcare
There is a wealth of open data available on healthcare costs, but can it be truly utilised? The more data the better you might think, but the usefulness of data greatly depends on its interconnectivity.
Bram, Christiaan, Koshin, Steven and Thijs investigated how healthcare data could be used to identify so-called hotspots. They condensed a vast amount of data into a dataset where connections were made between different data points. They then reduced this 'Swiss cheese' of data into a neatly filled Excel sheet, comparing a wide range of indicators per municipality. They used this data to test the usability of models by subjecting them to various machine learning approaches.
Smart vigilance zone
There are few countries in the world where, like in the Netherlands, most bicycle locks are worth more than the bike itself. Despite these robust chain locks, the Netherlands faces a bicycle theft problem, with an estimated 450,000 bikes stolen annually. The municipality of Utrecht is now investigating if IoT combined with RFID tags can be used to reduce bicycle theft. The concept works as follows: 2 RFID tags are attached to the bike and 1 RFID tag to the keychain. A UHF-RFID reader, similar to the gates at train stations but with higher power, is placed at a bike rack. When a person parks their bike, the reader registers the three RFID tags. After parking the bike, two transmitters (on the bike) remain while one transmitter (on the keychain) leaves the zone. The system then registers this as a person locking their bike and walking away. This data is then sent to the cloud via the LoRa network. If the two transmitters without the third transmitter leave the zone, it indicates that the bike is being stolen. However, there are still many questions surrounding this concept. What is the range of the UHF-RFID reader? Is LoRa a suitable technology for transmitting the data? How can the RFID tags be best attached to the bike? What are potential interference factors for reading the RFID tags?
To address these questions, Anton, Armand, Leon and Wessel set up a test installation at Hooghiemstraplein in Utrecht.
VRIs and bike data
Bicycle loops on the road are laid to communicate with traffic lights: when a cyclist rides over them, the traffic light 'knows' to turn green. A nice bonus, these loops can also be used to count the number of cyclists passing over them. It's not always accurate, as some cyclists may be missed. How can these loops be calibrated to measure accurately?
Daan, Jawad, Rami, Steven and Younes examined patterns in VRI data of cyclists in the city centre of Utrecht. They found that certain differences between loop counts are good predictors of 'busyness' (bike jams and congestion). This way, calibrating the loops for measuring the effects of certain infrastructure changes may not be immediately necessary.