Research shows that 75% of transport disruptions are unplanned, forcing operators to react slowly and inefficiently. This often results in a 20% increase in service costs, disrupted staff schedules, and a staggering 500 million hours wasted in waiting time.

This is a significant issue that needs immediate attention, and that’s where Moonbility’s solution comes in. Mobility aims to play a crucial role in helping the transport sector better manage disruptions. They address this problem with a platform that enabling operators to communicate disruptions with real-time updates more accurately and consistently, tailored to the specific stakeholders who need to be informed.

By doing so, they aim to empower people with reduced mobility, helping them regain confidence in using public transport. The solution not only enhances their trust in the system but also boosts their confidence, assuring them that they can complete their journey smoothly and without unnecessary stress. This sense of security and independence is key to creating a more inclusive and accessible travel experience for all.

Andre explains

“Moonbility’s mission is to help transport operators better communicate and manage disruption impacts.

For example, when there is a busy day, it is hard for bus operators to manage the service efficiently – our job is to make lives easier through real-time insights powered by our CCTV-based AI.”

What sets Moonbility apart is their unique approach to communicating disruptions across a network. Different stakeholders require different levels of information, and Moonbility tailor their system to meet those needs. For example, in a railway system, the operating team, maintenance crew, passenger information team, station control, and passengers all need varying degrees of detail about a disruption.

Moonbility leverage AI by introducing pre-built models that extract insights from existing data sets, such as CCTV footage and historical disruption data. Their machine learning algorithms analyse this data, and then apply a second layer of AI—large language models (LLMs)—to convert the insights into clear, human-readable messages. These messages are tailored to the specific stakeholders, ensuring that the maintenance team receives relevant, actionable updates in their preferred communication format. This streamlines communication and ensures that every team receives the right information to respond efficiently to disruptions.

Their system utilises onboard CCTV systems, applying machine learning and computer vision techniques to analyse the footage and extract insights regarding the occupancy of designated areas, such as wheelchair bays. These insights include detecting whether the bay is occupied by buggy users, wheelchair users, or individuals standing nearby. The CCTV systems process this data and send the insights via a real-time API for integration to existing systems. The resulting information is then forwarded to passenger-facing applications, depending on the operator’s specific requirements.

Andre tells us

“ What we are doing here is actually getting people to their destination quicker and more comfortably. So when people get rid of their private cars and start using public transport more and more, they know the choice is correct because that choice wouldn’t harm their travelling experience at all.”

The CAM Scale Up UK programme provides the opportunity to collaborate directly with test beds, enabling the Moonbility to work with Horiba Myra and Smart Mobility Living Lab, which has facilitated collaboration with three different bus operators. This enabled testing across different bus models, including buses in current operational use by TfL, allowing Moonbility to assess the latency and accuracy of their platform. The insights gained from these tests have validated the operation of the system in real-world scenarios, with problem free detection of wheelchairs, buggies, suitcases and people, and the data collected has helped further enhance the algorithms.

The value that the test beds have brought to the programme in addition to the real-life testing environment, is independent expert evaluation of solutions, particularly when integrated into existing transport infrastructure. This process helps the Moonbility team to identify system limitations, determine what works and what doesn’t, and assess potential challenges before implementation, serving as a critical step before mass deployment.

Andre explains:

“We’ve got this amazing environment for us to go around, move around, and actually to put our cameras in to test the limitations of our algorithms.

We’ve got an opportunity to test that our scenarios that are future-proof. For example, we tested cases where multiple buggies or multiple wheelchairs are on board at the same time and we’ve demonstrated our capabilities, and that the various objects are picked up without any issue.”

The strength of the technology lies in its model-agnostic nature, allowing it to operate on existing CCTV systems, provided the footage is available for analysis. This stage  of development has focused on analysing wheelchair usage by tracking the flow of objects in the footage, however during the course of the testing, the team identified a broader opportunity within the delivery sector, particularly in depot optimization. In large depots, operators move freely without the aid of technology to track their movements, making it difficult to improve operational efficiency. Moonbility addresses this challenge by enabling operators to extract valuable insights from existing CCTV data, allowing them to identify potential issues or opportunities and make informed decisions swiftly.

“A lot of people keep saying what we’re doing is very hard, the problem that we’re dealing here is almost impossible to solve. Thinking about the counterpart that we’re dealing with, the public, transport operators and also the government, given the nature of the beast, it is almost impossible for people to think about changing.

However, make that change and you are actually helping the entire society to have better experience than they could ever have imagined.”