The Digital Rail competition winner of one of the WM5G digital transport projects seeks to bring reliable streaming to rail and metro staff members in a bid to reduce safety and crime incidents on the system, and build public confidence back in public transport. Tangent sits down with Callum Gibson, Digital Rail Ltd, to find out more
What are some of the prominent difficulties in terms of accessibility and safety incidents such as trap-and-drag that you believe a CCTV 5G system feed would help alleviate?
CG: One of the main things is that we can send alerts to staff, but usually without 5G we won’t be able to send live video streams. I’m sure everyone has tried to use the internet onboard the tram or the train, and it’s never good. So, without 5G, we’d only be able to send images to staff if an incident does occur. With 5G, we can live stream with video, we can send everything we need and some surrounding information if something does happen.
There’s a lot of demanding stuff going on in terms of processing requirements. It needs to be the right hardware, with the right kind of power fitted to the train
In addition to that, the security aspect is quite interesting, because obviously there’s a lot going on – we’re detecting when someone moves a bag on the tram, or takes a bag from the tram – so there is a lot of processing has to be done. There’s a lot of demanding stuff going on in terms of processing requirements. It needs to be the right hardware, with the right kind of power fitted to the train. The more cameras there are, the more processing is needed.
You’ll be retrofitting existing CCTV systems capable of streaming 5G feeds to staff members’ devices. How does this project look in terms of costs, timescales, and the number of vehicles being piloted?
CG: It depends on what systems are in there. But if you assume that the cameras are on there already, it’s just a case of working with the provider of the CCTV and accessing their system. Most of the cameras are just IP cameras on the network, so as long as we can tap into the network, we can tap into the cameras.
In terms of costs, it’s low. I couldn’t give you a specific amount, and prices would vary from provider to provider. Some of them might be easier to tap into than others, but it depends on how it’s all wired up. But I can tell you that as far as things go, all we need to install is a box. We just need to install processing units which can just tap into the cameras that are already there. For ease of installation, we’re right at the top.
It’s not a large number – I think it might only be a couple of trams. We’re demonstrating the proof of concept, with the hope that, with a successful test, we can roll it out to more.
How will staff actually be notified if an incident occurs? Is there a priority system for who is notified?
CG: It will be in a dedicated app. They will get a notification, and then once they open the notification, they will be able to look at all of the information of the livestreams from the actual cameras that have generated the notification.
We’re demonstrating the proof of concept, with the hope that, with a successful test, we can roll it out to more
If you see for example someone with a wheelchair, that alert will have a snapshot, and then you’ll be able to open the video stream of the camera that picked up the wheelchair. So you can see the situation in real time.
It depends mostly on the use case. Basically, who can deal with it: for example, with trap-and-drag incidents, you want to tell the driver immediately. They’re the one who acts and stops the tram, whereas for things like a disabled person coming on where there isn’t any disabled access, you’re going to want to notify staff who can help them get on.
It’s important, in encouraging passengers back onto public transport following COVID-19, to make things as safe and secure as possible, isn’t it? How can this project help do that?
CG: Obviously, a lot of this is about public dissemination. One of the use cases that we’re going for as well is the passenger counting and business system. Particularly during the whole COVID situation, it had a big effect, because we can regulate the business of a carriage. If a particular carriage arrives, we have the ability to know how busy it is. And we can keep that average much lower when people have to social distance.
For just a nine-month project, what does success look like for a trial such as this?
CG: In terms of getting people to come on, it’s just about raising awareness about our system, about all of the stuff that’s going on. Especially for disabled access, if people with accessibility issues are now more aware that we’re keeping an eye on the platforms that they use – and if we find that there are a lot more wheelchair users that don’t have disabled access, then the right things can be done to make sure that they do have disabled access.
There’s a lot of action to be taken as a result of this that will genuinely improve everyone’s experience. We hope that the results are passed on back to our tech. The nine months has already begun, they began when we signed the contract.
Success really is just wanting to make a lot of detections. Obviously not false positives, but we’ll know that we’re doing well if we trigger a lot of alerts at the right time. A long part of this is the testing phase, out of the nine months, more than half is going to be on actual deployment, making refinements, because the technology largely is already developed. So, the point is to basically refine this as much as possible, do testing, do as many studies as we can, and invite a lot of stakeholders along that might be interested in actually demonstrating the technology.