4 minute read time.

What is blocking an integrated road transport system?

On the face of it, Transport would appear to be an excellent example of a large-scale engineering system, rather than a collection of disparate technologies that move people and goods from A to B.   The plain fact is that this is not true, not entirely.  For while cars run on roads, there is little actual integration to make it a true system (other than through the driver).

To begin with, Systems Engineering (SE) principles are not applied to this wider road transport system of systems, which includes the vehicles, road infrastructure and communications networks to connect them in the future. Indeed, there is no one single organisation or authority that has this system and data integration as a core goal. It is difficult to see how the anticipated future developments (such as autonomy) can be delivered effectively without an SE approach spanning the industries, services and government agencies involved.

An important question is ‘why SE approaches have not yet been applied to road transport systems at a system of systems level?’  The answers can be summarised in this way:

Lack of systems thinking in each individual domain The vehicle manufacturers focus on their domain and the roads and infrastructure organisations focus on their domain. There is limited collaboration. 

Unclear ownership of the overall road transport objectives including functionality. Vehicle manufacturers, roads and infrastructure organisations have no common vision of the long-term future of mobility.  They lack a clear overarching authority to push collaboration to meet high-level goals and objectives. As a consequence, they develop different solutions; as we have seen with the installation of vehicle electrification and charging framework. Here, a proliferation of means of accessing charging infrastructure (including connection technologies, standards, payment methods, user interfaces) have slowed roll out while creating a negative impact on user experience and public perception. 

Lack of standards and legislation There are few international standards for vehicle-to-infrastructure interoperability. Communication technologies and protocols is only one example. Such standards will become necessary as communication is needed to allow autonomous driving in complex environments such as cities. 

Competition in the automotive domain prevents collaboration on future functionality (which constrains generation of standards) Automotive manufacturers are unwilling to share their concepts for future functionality; something that could reduce their competitive advantage if they can bring the technology to market first. 

Source: Julian Hochgesang on Unsplash

Despite these challenges, there are opportunities within existing organisations. The deployment of an SE approach will inevitably require change. Many such changes are aligned with the experiences of practising engineers. For example, there is a significant commonality between SE approaches and project management best practice. Alongside this, it is common for many major projects to have a Systems Engineering Management Plan (SEMP), which defines the SE approaches to be used on the project. Many engineers involved in the delivery of transport systems have had exposure to a system architecture in some form (although they might not recognise it as such). 

The Green Light 

A good example of automotive and transport infrastructure integration involves the use of traffic light to vehicle communications. Green Light Optimised Speed Advisory (GLOSA) has already been trialled. It provides advice to the driver on when to slow down in order to pass through the traffic light on green or prepare to stop. In addition, with increased autonomy in vehicles, this type of communication could be used to adjust their speeds and/or route based on this real-time data. 

It could also:

  • improve traffic flow
  • reduce local pollution
  • reduce inconvenience to other road users
  • reduce the number of stationary vehicles. 

It isn’t clear if this vehicle to infrastructure support would specifically deliver on these desired outcomes. Nor whether these outcomes are the only potential benefits the system could deliver. In addition, with increased autonomy in vehicles it is suggested that this type of communication could be used to adjust their speeds and/or route based on this real-time data.

Vehicle-to-Infrastructure communication is often seen as a requirement or aid to deliver fully autonomous vehicles in an urban environment. If this were to be an aim of a system such as this, it would be wise to include requirements on the speed and accuracy of data from the system, failure analysis etc. In addition, this could cause significant customer satisfaction issues where autonomous features were made unavailable due to infrastructure problems. 

The automotive market is a highly competitive one, albeit with a highly shared supply chain. Thus, collaboration between manufacturers on agreeing standards for infrastructure interfaces is difficult to organise and agree; particularly when the benefits would be shared by all without offering a competitive advantage. Ownership of the problem and its solution is also a major challenge in this scenario.

Share your thoughts!

Who will manage data that transport operators and users rely on, and how can SE be used?

Is it possible to provide a properly integrated transport system without a national policy to drive it?

How can SE meet the needs of the voiceless (e.g. partially-sighted pedestrians) in an arena dominated by government and corporate entities?

Authors: Matthew Clarke, Stephen Powley, John Kelly, Iain Cunningham, Vanessa Mascall, Andy Harrison, Dr. Andrew Hussey, Gareth Topham, Dr. Raj Takhar, Dr. Michele Fiorini, Jana Skirnewskaja, Kareem Drysdale, IET Transport Panel Ecosystems Challenge Group. Partner organisation: INCOSE UK

#thewholesystem 

Parents
  • Regarding Green light operation-  It would be a good idea to give Emergency vehicles wireless access/control to signal lights.

  • Interesting point about the unintended consequences of the green wave. Definitely a thing that an SE approach ought to be considering in the trade-off analysis.

  • I remember being shown a device for automatically setting green phase back in the early 1990's when I was a young technical assistant with Siemens Plessey (who made a good proportion of the traffic light systems at that time). So the idea has indeed been around a very long time.

    I suspect cybersecurity is the last of their problems compared with retrofitting a system onto older traffic lights systems - I'm pretty sure when I've seen the boxes opened up near me, in some cases they look very much like the systems I worked with.

    The other problem is, if you set everything in a green wave, you may cause unexpected emergent properties which result in traffic slowing down even more and delaying the ambulances instead of helping. The phasing is usually modelled (from memory). This is an example of a properly integrated system in action - over a small geographic area at least and one specific type of traffic.

    As an aside (sorry, I promise to keep it on topic) - I recall one of the profitable services that Siemens had at the time was resetting the phasing because the local authorities had made a mess of it.

  • That does indeed sound like a good idea. I did a quick search and it's been an active area of research for at least the last decade (https://ieeexplore.ieee.org/document/10142622, https://www.indusedu.org/pdfs/IJREISS/IJREISS_4293_36892.pdf). There's at least one commercial solution (https://www.gtt.com/emergency-response/), but there are security challenges (https://www.thedrive.com/news/hacker-uncovers-how-to-turn-traffic-lights-green-with-flipper-zero).
    This definitely feels like a case where applying Systems Engineering to the whole system would be beneficial to the adoption of safe and secure technology for maximum benefit.

Comment Children
  • Interesting point about the unintended consequences of the green wave. Definitely a thing that an SE approach ought to be considering in the trade-off analysis.

  • I remember being shown a device for automatically setting green phase back in the early 1990's when I was a young technical assistant with Siemens Plessey (who made a good proportion of the traffic light systems at that time). So the idea has indeed been around a very long time.

    I suspect cybersecurity is the last of their problems compared with retrofitting a system onto older traffic lights systems - I'm pretty sure when I've seen the boxes opened up near me, in some cases they look very much like the systems I worked with.

    The other problem is, if you set everything in a green wave, you may cause unexpected emergent properties which result in traffic slowing down even more and delaying the ambulances instead of helping. The phasing is usually modelled (from memory). This is an example of a properly integrated system in action - over a small geographic area at least and one specific type of traffic.

    As an aside (sorry, I promise to keep it on topic) - I recall one of the profitable services that Siemens had at the time was resetting the phasing because the local authorities had made a mess of it.