Implementing and Iterating WeWALK’s Agent-Based Guidance System (WeASSIST) in Rail Transport to Improve Visually Impaired Customer Experience

英国有超过200万视力受损人士（vii） （AgeUK\_2015）；全球估计为2.53亿（Ackland\_et\_al\_2017）。随着人口老龄化，这将增加，因为视力丧失不成比例地影响老年人。vii用户可以借助白色手杖四处走动，通常还可以配合智能手机导航应用。然而，这只能检测到距离用户大约1米的膝盖以下的障碍物。大多数智能手机的界面都是不可访问的，这限制了它们的实用性。这使得vii在繁忙的城市地区安全导航变得具有挑战性，特别是在有限的gps可用性下需要关键任务导航和危险意识的铁路环境。为了应对这一挑战，WeWALK开发了一种革命性的“智能”设备，可以固定在用户现有的手杖上。它可以检测到膝盖到头部高度的障碍物，并通过与谷歌地图、Moovit和微软服务的集成，独特地连接到用户的智能手机，提供基于音频的导航。为了增强WeWALK的室内功能，我们创建了一项新的服务——WeASSIST。WeASSIST使用智能手机的摄像头，通过视频馈送将vii与训练有素的代理连接起来。在WeWALK的定位技术和易用性的辅助下，代理可以在任何环境中引导用户，根据用户的要求提供个性化的描述。该项目将在铁路环境中实施WeASSIST，解决vii无法定位目的地/运输路线的最后一英里问题，创建一个全面、安全关键且可访问的导航系统。WeASSIST正在不断发展，可以将计算机视觉集成到物体识别和映射中。然而，这还没有被证明具有足够的系统性能（准确性/可靠性/可用性）的铁路应用。与RNIB合作，我们将解决技术/安全方面的挑战，以防止在铁路环境中广泛采用基于智能体的导航系统。我们将以现有的WeWALK/WeASSIST架构为基础，使用新软件为铁路运营商/客户提供安全关键功能。该项目可以发挥至关重要的作用，确保越来越多的视障人士能够充分利用铁路旅行，同时改善工作人员的可用性和客户体验。

There are over 2-million Visually Impaired Individuals(VIIs) in the UK(AgeUK\_2015); globally estimated at 253-million(Ackland\_et\_al\_2017). As populations age, this will increase as sight-loss disproportionally impacts older people. VIIs benefit from the white cane to get around, often in conjunction with smartphone navigation apps. However, this only detects obstacles below knee-height within approximately 1m of the user. Most smartphone interfaces are inaccessible, limiting their usefulness. This makes it challenging for VIIs to safely navigate busy urban areas, particularly railway environments requiring mission-critical navigation and hazard-awareness in limited GPS-availability.To address this challenge, WeWALK have developed a revolutionary "smart" device that screws onto a user's existing cane. It detects obstacles at knee-to-head height and uniquely connects to a user's smartphone to provide audio-based navigation via integration with Google Maps, Moovit, and Microsoft services. Efforts to enhance WeWALK's indoor capabilities resulted in the creation of a new service, WeASSIST. Using a smartphone's camera, WeASSIST connects VIIs with trained agents through video-feed. Supplemented by WeWALK's positioning technology and ease-of-use, agents guide users through any environment, providing personalised descriptions at the user's request. This project will implement WeASSIST in a railway environment, solving last-mile problems where VIIs fail to locate their destination/transport route, creating a comprehensive, safety-critical, and accessible navigation system.WeASSIST is in constant development and can integrate computer-vision for object-recognition and mapping. However, this has not yet been demonstrated with adequate system performance (accuracy/reliability/usability) for rail applications.Partnering with RNIB, we will address the technical/safety challenges preventing widespread uptake of agent-based guidance systems in railway environments. We will build on existing WeWALK/WeASSIST architectures with new software to provide safety-critical functionality to railway operators/customers. This project can play a vital role in ensuring the rising numbers of visually impaired people have full access to rail travel while improving staff availability and customer experience.