James Day speaks with Simon Tong, Head of CAV Research and Deployment at TRL, along with Heathrow Airport, to discover more about the journey of autonomous pods through the Greenwich GATEway project
One of the most high-profile autonomous vehicle projects in the UK is Greenwich GATEway. The project is currently gearing up for the first UK trial of the pods on roads where they will share space with other road users, whether that be pedestrians, cyclists or vehicles.
Ahead of the latest stage of the trial, which is expected to begin this month, I was able to speak with the team at Heathrow, who first introduced vehicles of this nature at the airport’s Terminal 5 back in 2011, along with Simon Tong, Head of CAV Research and Deployment at the Transport Research Laboratory (TRL), who is heavily involved in the project, to find out more about how the technology has been progressing.[wlm_nonmember][…]
By subscribing you will benefit from:
- Operator & Supplier Profiles
- Face-to-Face Interviews
- Lastest News
- Test Drives and Reviews
- Legal Updates
- Route Focus
- Industry Insider Opinions
- Passenger Perspective
- Vehicle Launches
- and much more!
Beginnings
The GATEway project started after a consortium bid into Innovate UK, when it put out the first call for projects to explore connected and automated vehicles within the UK.
Innovate UK funded three projects, with GATEway being the Greenwich and London-focused project. The others are Autodrive in Milton Keynes, the other Ventura in Bristol.
All have different approaches, with the key focus of the GATEway Project being that it wanted to demonstrate first and last mile transport using driverless vehicles, ideally using some kind of shuttle of pod. The Greenwich peninsula was identified as a good location, because of key transport hubs in the north (Underground, Thames Clipper ferry and the Emirates Air Line cable car) along with a lot of businesses and residences further south and further down into the borough without good transport connectivity. It was seen as a good example of where driverless pods could be used to create a compelling demonstration of what last mile transport might look like using pods.
The vehicle consortium brought together Westfield Sportscars, who would build the pods using a Heathrow design in use at Heathrow Terminal 5, and a Bristol-based company called Fusion Processing which is implementing the systems which make all of the pods driverless.
Simon explained: “The key difference with the pods at Heathrow is that although they run on wheels and are effectively driverless, all the autonomy and connected systems are built into the concrete guideway which they run in.
“With the GATEway project, we asked how we could get those autonomous pods running without all that supporting infrastructure installed. You would then have the capability of using them in a variety of locations without having to build something specifically.
“We took that design, which we knew was functional and reliable in terms of providing a transport service, and retrofitted all of the kit on it, which will make it driverless and enable it to work in different areas.”
Specification
The vehicles are fully electric, running with a lithium phosphate battery pack. The electronic control system is a unique Heathrow design put together for the Heathrow application, providing all the drive and basic control system of the vehicle.
The Fusion system has a number of sensors which are tailored for different distances. They use radar, cameras and lidar for medium-to-long distance obstacle detection, and they have ultrasound sensors around the vehicle which provide proximity detection for up to about five metres.
On top of that, the pods use highly accurate GPS systems. A GLSS module on the vehicle is the basic navigation tool, while additional systems on the vehicle help with localisation if the GPS signal is not quite accurate enough in certain areas, such as around tall buildings.
“The unique thing about Fusion is that all of their sensors are, to a certain extent, designed and developed in, house, rather than being off-the-shelf products,” said Simon.
“What they do is focus heavily on having an efficient system. A lot of driverless vehicles have been shown by other companies to have very high rates of data collection from high resolution sensors, which are fed into one or more massive supercomputers within the vehicle.
“Fusion tries to operate in a more efficient way, which is potentially much cheaper and easier to commercialise, by having a lot of the data processing done directly at the sensor head. This means they don’t have a massive supercomputer onboard, but a small ECU which receives inputs from all of the different sensors where typically a lot of the processing has already been done. The ECU then provides control inputs to the vehicle.”
The multitude of different systems helps the vehicles to operate effectively in a variety of conditions and anticipate obstacles at various distances.
“We know from some of the other work we did earlier in the project, using just lidar and camera-based systems, that sensors can be susceptible to different lighting conditions and don’t necessarily operate well in poor weather,” Simon added. “The benefit of the fusion system is that using additional technology like ultrasound and radar, you can get past some of those environmental limitations.
“We expect our pods to operate well in a variety of weather conditions, though we haven’t had the opportunity to test this yet. With trials running into winter, I’m sure we’ll have the opportunity to see this in practice soon enough.”
Public trials and perception
A public trial of the Greenwich GATEway autonomous pods is expected to launch sometimes this month. When I spoke to Simon, the first production pod was undergoing final preparations, checks and validation, with the second pod on the way. The full fleet used in the trial will consist of four vehicles overall.
Not all of the work undertaken by the project so far has focused on passengers. One of the other aspects of the project is last mile urban deliveries, which used a vehicle called Cargopod in June.
“The single vehicle was a small electric city delivery vehicles, with the load pay adapted and configured so that it would work as a driverless grocery delivery vehicle,” Simon explained.
“The trial lasted for a couple of weeks and delivered to over 100 different customers in a residential area in Greenwich. It was the first time a service like this has been demonstrated in the UK.”
The project has also examined public perception and understanding of driverless vehicles, both before and after people had experienced the vehicles on the road.
“Overall, the public perception has started very positively,” Simon remarked. “It has remained that way based on the surveys we have done so far.
“With the specific trials we’ve done with members of the public, the overall message has been positive and people are starting to realise that driverless vehicles can perform a range of functions and they don’t just have to be driverless versions of conventional cars. This is how a lot of members of the public tend to see them initially.
“With GATEway, people are seeing the potential for public service and ridesharing models, which has been quite an enlightening experience for them. They appreciate that this could revolutionise the way environments are laid out and people get around.”
Infrastructure
Simon said that the connected side of connected and automated vehicles needs more focus before vehicles of this nature become more widespread: “Companies are definitely getting the hang of how to make vehicles driverless and carry out basic driverless functions, but it’s that network optimisation which is the really big step that needs to be addressed.
“It’s all very well having a driverless vehicle, but you then need to make it act like other human driven vehicles do, or at least interact with them and traffic signalling and networks, so that they can move efficiently and integrate with existing road networks. The connectivity will really help with that.”
Bullying
One concern about autonomous vehicles is the possibility of other people bullying the vehicles, whether that be other road users pulling out in front of them and forcing them to stop, or pedestrians maliciously preventing the vehicles from moving around. I asked whether in the pedestrian-heavy areas of Greenwich where the vehicles had been demonstrated, this was a concern.
“It is certainly something we have seen evidence of,” Simon replied, “but I think it’s more out of the novelty rather than wanting to gain some sort of advantage over the pod in terms of using the space.
“When people see them for the first time, they do sometimes step in front of them, because they want to see how they respond to obstacles ahead of them. But when we ran the prototype in April for just under a month, that behaviour quickly disappeared. Pedestrians become quite familiar with the vehicles and tend to let them operate in their own space.
“From the work we did in our driving simulator, it was not a finding that came out. However, the comments people left suggested that if people wanted to take advantage of driverless vehicles, they would do so by pulling out in front of them and leaving quite small gaps, assuming they would stop quite quickly.”
Monetising
While the vehicles currently have a supervisor onboard, the intention is that they would eventually operate without needing any staff member travelling with them to ensure they run safely. This would in turn mean that nobody is on board to collect fares.
“I’d say the intention would be for the vehicles to operate in similar ways to other modes of transport – perhaps with contactless payments providing access to the vehicle,” Simon suggested.
“There’s certainly things that could be done to ensure payment was collected.”
Congestion
Does Simon foresee the vehicles easing congestion?
“They definitely have potential to relieve congestion if they can take people away from using vehicle transport on roads, or at least reduce the number of vehicles being used. Vehicles do need to be integrated effectively into the existing road network.
“If you get local authorities on board and plan where services will go so they do provide the connectivity people need to make better use of public transport, rather than replace it, then I really feel the could impact positively on congestion. It will require cooperation from authorities like TfL to make sure you aren’t simply providing alternatives to already effective public transport, like the tube, and putting vehicles back on the streets instead. I don’t think that’s they way they should be used.
“Driverless vehicles have great potential to match the demand that there is for a particular service at a particular time, not using a larger vehicle where a smaller vehicle would do.”[/wlm_ismember]
