BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//talks.staging.osgeo.org//foss4g-2024-academic-track//spe
 aker//CU8MFT
BEGIN:VTIMEZONE
TZID:-03
BEGIN:STANDARD
DTSTART:20000101T000000
RRULE:FREQ=YEARLY;BYMONTH=1
TZNAME:-03
TZOFFSETFROM:-0300
TZOFFSETTO:-0300
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-foss4g-2024-academic-track-XMAGHX@talks.staging.osgeo.org
DTSTART;TZID=-03:20241204T174500
DTEND;TZID=-03:20241204T181500
DESCRIPTION:## 1. Introduction\nUrban transportation is transforming with a
  focus on sustainability and smart city initiatives. Cycling\, a key eleme
 nt of sustainable urban mobility\, needs robust infrastructure and reliabl
 e data for growth and integration into city planning. Despite advancements
  in sensor technology and (geo)data analytics\, there is a gap in comprehe
 nsive collection and use of cycling-specific environmental\, safety\, and 
 pathway data.\n\nOne major deterrent for citizens using bicycles are the p
 erceived dangers in traffic. Identifying insecure sections is crucial to i
 mproving cycling infrastructure. Safe countermeasures can change negative 
 perceptions and promote cycling as a safe and sustainable mode of transpor
 t. Traditionally\, only actual crashes are included in official data\, inf
 orming city planning decisions. However\, analysing high-risk occurrences 
 like near-miss incidents\, which greatly impact the perceived danger\, can
  provide a more accurate understanding of cycling safety.\n\nThere already
  exist a number of projects using different technologies to gather and pro
 vide data on bicycle safety and urban mobility\, but combining environment
 al and road safety aspects is unique. Projects examining cyclist safety\, 
 particularly dangerously close overtaking manoeuvres\, often involve remot
 e data processing with machine learning or human analysis. Using live vide
 o or images from bike-mounted smartphones is effective but creates data ov
 erhead and privacy concerns. Additionally\, microcontroller-based sensing 
 systems can be complex to assemble\, requiring technical skills and specia
 l equipment.\n\nThe objective of this work is to address the aforementione
 d gap by developing an innovative bicycle sensor system that leverages emb
 edded artificial intelligence (AI) to process sensor data on the device. T
 his approach has the potential to reduce data overhead and address privacy
  concerns while simultaneously providing actionable insights. Our work has
  the potential to make significant contributions to traffic and transport 
 planning by providing valuable insights into traffic patterns and road saf
 ety concerns using extensive spatial datasets gathered by citizens.\n\n## 
 System Design\nAt the core of our system is a microcontroller unit (MCU) o
 f the senseBox family. The senseBox is a versatile\, open hardware electro
 nics kit specifically designed for citizen science projects and educationa
 l initiatives\, with an emphasis on environmental monitoring and data coll
 ection.\n\nThe following environmental sensors are used:\n- Temperature & 
 rel. Humidity (HDC1080)\n- Particulate Matter (SPS30)\n- Acceleration (MPU
 6050)\n- Time-of-Flight (ToF) ranging (VL53L8CX)\n\nMoreover\, battery man
 agement\, Bluetooth Low Energy (BLE)\, and OLED-Display modules are includ
 ed for connectivity and user feedback. All parts fit into a custom designe
 d\, 3D printed enclosure which is attached to the seat post of a bicycle.\
 n\nThe device is communicating with an open source smartphone app using BL
 E which receives sensor data and combines them with geolocation data. Data
 sets are recorded and saved on the smartphone\, but can also be uploaded t
 o openSenseMap as open data during the ride. Users can control levels of p
 rivacy (e.g. by setting privacy zones) to foster digital sovereignty. \n\n
 ## 2.1. Machine Learning on the Bike\nWe are introducing two approaches to
  utilise machine learning capabilities using Tensorflow Lite on the sensor
  device: overtaking detection and road surface / quality classification. B
 y processing the data directly on the device instead of sending it to larg
 er servers\, bandwidth and energy consumption is kept minimal.\nIn the low
  resolution depth images recorded by the 8x8 multizone ranging ToF sensor\
 , overtaking vehicles can be detected using shallow neural networks. This 
 has already been described and implemented as a standalone solution in (Sc
 harf et al.\, 2024)\, but for integrating it into the mobile sensor system
  some considerations for available processing capacities\, suitable infere
 nce times and necessary accuracies will be addressed as part of this work.
  \nTo classify the road surface and its quality\, the acceleration sensor 
 will be used. While raw acceleration values can identify the roughness of 
 a road\, surface classifications and quality estimations can reveal deviat
 ions from intended surfaces to actual surfaces. Using acceleration values 
 and geolocation data\, we will explore training a machine learning model u
 sing OpenStreetMap Surface information as ground truth data.\n\n## 3. Work
 shops\nEngaging citizens in data collection\, problem identification\, and
  the construction of sensor stations empowers them and fosters the generat
 ion of new ideas. Our solution is a solder-free\, easy-to-assemble mobile 
 sensor device. We conduct a workshop in São Paulo\, Brazil\, where 20 par
 ticipants build and mount their own mobile sensor device on bicycles. Afte
 rwards\, they collect environmental and bicycle-specific sensor data. Foll
 ow-up workshops in Münster\, Germany will allow the comparison of the con
 trasting bicycle infrastructures in these cities\, as well as the general 
 urban environment differences\, and will provide valuable data and insight
 s into participants' perceptions.\nThis collaborative effort enhances part
 icipants' understanding of scientific methods and urban mobility challenge
 s while ensuring that the collected data reflects cyclists' authentic expe
 riences. By involving citizens as active contributors\, we aim to bridge t
 he gap between scientific research and community needs\, fostering a more 
 inclusive and participatory approach to urban mobility solutions.\nAfter t
 he workshops we conduct user studies with the workshop participants on the
  following topics:\nUsability: Through surveys at the end of each session 
 and interviews\, participants provide feedback on assembling\, mounting an
 d connecting the bicycle sensor device. \nTrust in Data: Participants revi
 ew the data of their recorded dangerous takeovers and road surface types a
 nd compare its accuracy with their own perceptions.\n\n## 4. Conclusion an
 d Future Work\nThis comprehensive evaluation aims to provide a thorough un
 derstanding of both the user experience and the technical performance of t
 he system\, ultimately guiding the data-driven foundation for improvements
  in urban mobility solutions. Insights gained from this work will inform f
 uture iterations of the project\, ensuring the system collects high-qualit
 y data and meets the needs of cyclists\, thereby effectively enhancing urb
 an mobility and road safety not only for cyclists but for all users of the
  urban mobility system. Future works will include the development of an op
 en source bike-related data analysis platform as a recommender system for 
 bike infrastructure measures in cities.
DTSTAMP:20260513T083943Z
LOCATION:Room I
SUMMARY:Urban Cycling: Intelligent Bicycle Sensors for Road Safety and Sust
 ainability - Felix Erdmann\, Luis Fernando Villaça Meyer\, Beatriz Gonça
 lves
URL:https://talks.staging.osgeo.org/foss4g-2024-academic-track/talk/XMAGHX/
END:VEVENT
END:VCALENDAR