Citizen Science for Traffic Monitoring: Investigating the Potentials for Complementing Traffic Counters with Crowdsourced Data

Traffic counts are among the most frequently employed data to assess the traffic patterns and key performance indicators of next generation sustainable cities. Automatised counting is often based on conventional traffic monitoring systems such as inductive loop counters (ILCs). These are costly to install, maintain, and support. In this paper, we investigate the possibilities to complement and potentially replace the existing traffic monitoring infrastructure with crowdsourcing solutions. More precisely, we investigate the capabilities to predict the ILC-obtained data using Telraam counters, low-cost camera counters voluntarily employed by citizens and freely accessible by the general public. In this context, we apply different exploratory data analysis approaches and demonstrate a regression procedure with a selected set of regression models. The presented analysis is demonstrated on different urban and highway road segments in Slovenia. Our results show that the data obtained from low-cost and easily accessible counters can be used to replace the existing traffic monitoring infrastructure in different scenarios. These results confirm the prospective to directly apply the citizen engagement in the process of planning and maintaining sustainable future cities.

Quantifying changes in bicycle volumes using crowdsourced data

Most cities in the United States lack comprehensive or connected bicycle infrastructure; therefore, inexpensive and easy-to-implement solutions for connecting existing bicycle infrastructure are increasingly being employed. Signage is one of the promising solutions. However, the necessary data for evaluating its effect on cycling ridership is lacking. To overcome this challenge, this study tests the potential of using readily-available crowdsourced data in concert with machine-learning methods to provide insight into signage intervention effectiveness. We do this by assessing a natural experiment to identify the potential effects of adding or replacing signage within existing bicycle infrastructure in 2019 in the city of Omaha, Nebraska. Specifically, we first visually compare cycling traffic changes in 2019 to those from the previous two years (2017–2018) using data extracted from the Strava fitness app. Then, we use a new three-step machine-learning approach to quantify the impact of signage while controlling for weather, demographics, and street characteristics. The steps are as follows: Step 1 (modeling and validation) build and train a model from the available 2017 crowdsourced data (i.e., Strava, Census, and weather) that accurately predicts the cycling traffic data for any street within the study area in 2018; Step 2 (prediction) use the model from Step 1 to predict bicycle traffic in 2019 while assuming new signage was not added; Step 3 (impact evaluation) use the difference in prediction from actual traffic in 2019 as evidence of the likely impact of signage. While our work does not demonstrate causality, it does demonstrate an inexpensive method, using readily-available data, to identify changing trends in bicycling over the same time that new infrastructure investments are being added.

Active Learning Based on Crowdsourced Data

The paper proposes a crowdsourcing-based approach for annotated data acquisition and means to support Active Learning training approach. In the proposed solution, aimed at data engineers, the knowledge of the crowd serves as an oracle that is able to judge whether the given sample is informative or not. The proposed solution reduces the amount of work needed to annotate large sets of data. Furthermore, it allows a perpetual increase in the trained network quality by the inclusion of new samples, gathered after network deployment. The paper also discusses means of limiting network training times, especially in the post-deployment stage, where the size of the training set can increase dramatically. This is done by the introduction of the fourth set composed of samples gather during network actual usage.

Monitoring Human-Wildlife Interactions in National Parks with Crowdsourced Data and Deep Learning

This short paper summarizes the first research stage for applying deep learning techniques to capture human-wildlife interactions in national parks from crowd-sourced data. The results from objection detection, image captioning, and distance calculation are reported. We were able to categorize animal types, summarize visitor behaviors in the pictures, and calculate distances between visitors and animals with different levels of accuracy. Future development will focus on getting more training data and field experiments to collect ground truth on animal types and distances to animals. More in-depth manual coding is needed to categorize visitor behavior into acceptable and unacceptable ones.

A Bayesian Approach to Estimate the Spatial Distribution of Crowdsourced Radiation Measurements around Fukushima

Citizen-led movements producing spatio-temporal big data are potential sources of useful information during hazards. Yet, the sampling of crowdsourced data is often opportunistic and the statistical variations in the datasets are not typically assessed. There is a scientific need to understand the characteristics and geostatistical variability of big spatial data from these diverse sources if they are to be used for decision making. Crowdsourced radiation measurements can be visualized as raw, often overlapping, points or processed for an aggregated comparison with traditional sources to confirm patterns of elevated radiation levels. However, crowdsourced data from citizen-led projects do not typically use a spatial sampling method so classical geostatistical techniques may not seamlessly be applied. Standard aggregation and interpolation methods were adapted to represent variance, sampling patterns, and the reliability of modeled trends. Finally, a Bayesian approach was used to model the spatial distribution of crowdsourced radiation measurements around Fukushima and quantify uncertainty introduced by the spatial data characteristics. Bayesian kriging of the crowdsourced data captures hotspots and the probabilistic approach could provide timely contextualized information that can improve situational awareness during hazards. This paper calls for the development of methods and metrics to clearly communicate spatial uncertainty by evaluating data characteristics, representing observational gaps and model error, and providing probabilistic outputs for decision making.

Where to hike? Mobile application for guiding tourists in the Bükkalja Region, Hungary

Hiking is one of the most popular outdoor sports activities in Hungary. Despite not having many mountainous areas, a wide network of hiking trails crosses the country’s landscapes. As online tourist maps and thematic mobile applications become more and more popular among hikers, the role of paper-based, analogue tourist maps decreases. However, no thematic application has been issued that contains detailed surveyed (or crowdsourced) data on attractions or the natural circumstances (coverage, difficulty) for a certain area in Hungary yet. Nature tourism in the Bükkalja Region, Hungary is mostly based on geological-geomorphological features that are completed with cultural facilities. The length of the hiking trail system is more than 370 km in the examined 354 km2 large sample area. We have developed an OS mobile application that offers guidance for tourists based on four basic pillars: the physical condition of the trails, the attractions along a trail, dangerous trail segments and hiking trail marking quality. These pillars are visualized with an OpenLayers-based online map. The result is a multi-purpose smartphone application. Its main aim is to offer a planning platform for tourists by examining the difficulty of the trails and designating the attractions to visit. There is information on the most important attractions of the area: cultural and geoscientific sites are also presented. We also encourage users to report changes to the map data content via the crowdsourcing menu. These comments and remarks are continuously checked for validity and the database is modified with the use of them.

Developing Information Technologies to Promote Dementia e-Friendly Communities for COVID-19 and Beyond

People living with dementia (PLWD) and their caregivers often face barriers to education, support, and services that can improve their health and quality of life. Information technology (IT) has been suggested as a solution to overcoming such barriers, though the development of evidence-based IT for dementia care is still developing. This project gathered stakeholder (e.g., providers, caregivers) perspectives on the development of a proposed IT solution to support community asset mapping that would allow families to self-assess their dementia-related service needs, educate them about available services, and link them with services they need in their community. This proposed IT would create a dementia resource database that relies on crowdsourced data from community stakeholders as well as relevant data mined from existing sources (e.g., CMS certified nursing home data). As part of the planning process, this project conducted qualitative interviews with providers and caregivers in four metro areas in Alabama and their surrounding rural communities to learn more about the content and features that stakeholders perceive as being most effective for the proposed technology. Stakeholders also discussed their experience of utilizing IT solutions during the COVID-19 pandemic to promote access and continuum of care when barriers to service intensified. Thematic findings provide detail on: 1) motivating factors among stakeholders to contribute crowdsourced data that support community members affected by dementia; 2) potential barriers to implementing IT for dementia support, based on experiences with IT use during COVID-19; and 3) how stakeholders envision IT to better connect community members with needed services.