SBS Transit turns to AI to tackle bus bunching, delays

SBS Transit is trialing an artificial intelligence (AI)-powered tool to help bus service controllers better manage delays and bus bunching, with the aim of improving service reliability across Singapore.

A demonstration of this tool, FlowOS, was held at Seletar Bus Depot on Thursday (Sept 24), giving the media a preview at how this technology can help public bus services operate more reliably.

When traffic congestion, bad weather or other disruptions delay buses, commuters may find themselves waiting longer than expected, only to see two buses on the same service arrive minutes apart, said SBS Transit in a press release on Thursday.

Behind the scenes, service controllers have to monitor buses in real time and decide when intervention is needed to keep services running smoothly.

But with each controller overseeing up to 70 buses at a time, tracking rapidly changing conditions can be a challenge.

According to the release, FlowOS analyses live bus operations, flags issues that require attention and recommends interventions to help maintain more even spacing between buses along their routes.

Tool doesn't replace service controllers: SBS Transit

The AI tool, said the operator, does not replace service controllers.

Instead, it draws their attention to potential issues and suggests possible responses, while controllers use their experience and judgement to decide on the best course of action.

SBS Transit Senior Service Controller Calvin Chan said: "FlowOS is like having a brilliant assistant next to me. It draws my attention to issues that may need action and suggests what I can do."

He added that he would then use his experience and judgement to decide on the best response for the service.

SBS Transit Group CEO Jeffrey Sim said factors such as adverse weather, accidents and roads being rendered impassable can affect buses and are beyond the operator's control.

"By employing AI technology through the FlowOS tool, we are empowering our service controllers to make better-informed decisions and respond more effectively to deliver reliable journeys for our commuters despite these factors," he said.

FlowOS is currently being trialled on Services 70 and 145.

The trial follows a Memorandum of Understanding signed between SBS Transit and UK-based Prospective Labs earlier this year to explore the use of FlowOS in local bus operations.

The technology has already been deployed by SBS Transit's sister company Metroline to improve bus service reliability in the United Kingdom.

helmy.saat@asiaone.com

新捷运正在试用一款人工智能 (AI) 工具，以帮助公交服务控制人员更好地管理延误和公交车扎堆情况，旨在提高新加坡各地的公交服务可靠性。

周四（9月24日），在实里达巴士车厂举行了FlowOS工具的演示，让媒体提前了解了这项技术如何帮助公共巴士服务更可靠地运营。

新捷运（SBS Transit）周四在一份新闻稿中表示，当交通拥堵、恶劣天气或其他干扰导致公交车延误时，通勤者可能会发现自己等待的时间比预期更长，却发现同一线路的两辆公交车会在几分钟内先后到达。

在幕后，服务控制人员必须实时监控公交车，并决定何时需要干预以保持服务顺利运行。

但由于每个控制器一次最多要监管 70 辆公交车，跟踪快速变化的情况可能是一个挑战。

根据新闻稿，FlowOS 分析实时公交运营情况，标记需要关注的问题，并提出干预措施，以帮助保持公交车沿途路线之间更均匀的间距。

工具不能取代服务控制器：SBS Transit

该运营商表示，人工智能工具并不会取代服务控制人员。

相反，它会引导他们注意潜在的问题并提出可能的应对措施，而控制人员则运用他们的经验和判断来决定最佳的行动方案。

新捷运高级服务总监陈嘉文表示：“FlowOS就像我身边有一位出色的助手。它能让我注意到可能需要采取行动的问题，并提出我可以采取的措施。”

他还补充说，他将运用自己的经验和判断力来决定对该服务的最佳回应。

新捷运集团首席执行官沈志伟表示，恶劣天气、事故和道路无法通行等因素都会影响公交车的运营，而这些因素超出了运营商的控制范围。

他说：“通过 FlowOS 工具运用人工智能技术，我们能够帮助服务控制人员做出更明智的决策，并更有效地应对这些因素，从而为我们的通勤者提供可靠的旅程。”

FlowOS 目前正在 70 号和 145 号服务上进行试用。

此次试验是继 SBS Transit 与英国 Prospective Labs 今年早些时候签署谅解备忘录，探索在本地公交运营中使用 FlowOS 之后进行的。

SBS Transit 的姊妹公司 Metroline 已经部署了这项技术，以提高英国的公交服务可靠性。

helmy.saat@asiaone.com