NetworkSim shows visually and with data outputs how a network will perform with different infrastructure or timetable options. It takes industry standard inputs (in rail from timetable tools like RailSys, Hastus or Austrics) or IoT/logistics networks time stamps. We build a schematic of your physical network and set vehicles to interact with the network based on your inputs (either real-life timestamp data, or planning tools like GTFS timetables).
NetworkSim then statistically evaluates the quality of different infrastructure options or timetables by simulating the results. This is particularly powerful where there are too many options to check manually, or when legacy timetable tools take too long to build and change (or don't have enough model runs to be statistically valid).
NetworkSim has a number of modules that can be used on a stand-alone basis or as part of a wider deployment as follows:
NetworkSim has a dwell time module to help calculate the most likely dwell times for your public transport network. Our model is built on the basis of the Boston Dwell Time model in academic research and blends expected patronage with systems performance to calculate the necessary dwell times. The model builds a dwell time distribution that NetworkSim can draw from in its core product, or it can be used as a stand-alone module.
Our DRT module is designed to simulate performance of your demand-response transport services. Traditional timetable software requires fixed routes and fixed journey times to properly estimate service performance (including waiting time, in-vehicle time and journey time), vehicle kilometres and staffing requirements. NetworkSim overcomes this by randomly generating tens of thousands of pick-up requests within a proposed on-demand zone. Demand parameters can be set so they exclude areas with traditional public transport, or ignore areas with limited population density. The resulting distribution of journey times form a robust estimate of vehicle service hours, kilometres and performance. We include live traffic journey time requests, so the actual matches the plan. The zones can also be backwards-engineered to match specific service times (e.g. what size would the on-demand zone need to be in order to provide a 10-minute pick-up guarantee with 2 vehicles running).
Building a dynamic simulation model.
NetworkSim creates an infrastructure schematic based on your network data (physical network structure, including timing points, entry/exit rules, rail signalling rules, stops, loading bays or termini and any other relevant requirements).
We then build a 'timetable' either by taking GTFS format from your timetable systems, or by building headway/scheduling models, or by using your real-world distributions of time to plot how vehicles actually travel through your network. The bigger the data set the better (so we can plot a full statistically valid distribution).
We build in adjustments to the network for you to test. This can be loading bay rules for logistics networks, or different infrastructure options for rail networks (e.g. different platforms, crossover or station configurations or driver behaviour).
We then run NetworkSim thousands of times until we get a stable, statistically valid representation of your network. We can compare the relative performance of each infrastructure, timetable or other option, and encode your KPIs if you need to calculate the consequences like financial penalties of delays or disruptions to your network.
This is in a nutshell the ❤ of the NetworkSim.
NetworkSim contains multiple features that individually can be used to highlight and strengthen analysis.
Specify how you'll run your network - we're flexible enough to work with any operation type from random turn-up, to fixed timetable, to headway services.
We use industry standard inputs (e.g. GTFS timetables) and have worked with all the major planning tools (e.g. Hastus, RailSys, Austrics, VISSIM). Custom input is also supported.
Find your risks, test what happens to your network when things go wrong. Build, test and evaluate contingency plans to suit your specific requirements.
Do you have complex networks, with independent distributions of delays for each part? No problem, NetworkSim uses the individual component distributions and statistically combines them together. This is especially important for blended networks.
Test unlimited numbers of infrastructure options. Test different truck loading bay configurations, or platforms for stations to produce the best results for your operation.
What did that delay cost you? Build a payment mechanism into the tool to dynamically price the delays on your network and get full insight into your costs.
See below how NetworkSim is being used in different settings to demonstrate the power of simulation.
To find an optimal network infrastructure, timetable and risk profile NetworkSim was used for a new light rail network. The available data from industry standard timetabling and traffic light tools was used and loaded into NetworkSim. Next, multiple network configurations, timetables and driver behaviour profiles were evaluated and converted into objective insight into performance and associated costs. This enabled the executive team to perfectly tune their bid details with their desired risk appetite and simultaneously demonstrate this in a transparent and intuitive way.
NetworkSim can provide an objective data driven simulation of how your network will perform under different timetables or operational conditions, improving how you schedule train platforming or junction crossings. This is especially important to measure how different timetables will impact your operational performance and the associated costs or to build robust contingency plans - choosing the best way to recover from degraded modes.
NetworkSim can produce the expected performance of your chosen operating rules - allowing you to bid the best price with the most competitive timetable (without breaching KPI thresholds) for your risk appetite. NetworkSim can give you a competitive advantage over your competitors who have to rely on legacy timetable tools with limited or no dynamic simulation capability.
NetworkSim can determine the most efficient loading rules for trucks at your terminal - should you streamline loading by product type, or by truck type? Or a blend of both. NetworkSim can test all of the possibilities without having to actually change any of the infrastructure. It takes the trial-and-error out of network planning optimisation.
How do you make sure that your timetables can actually be delivered, or you're not paying for performance padding? NetworkSim can ensure that you understand what you're asking of your operators, and test their timetables before you approve them. Evaluate a payment mechanism and test if it produces the necessary results.
Mobility affects us all and is one of the key drivers for an efficient economy. A technological revolution is causing the industry to rapidly change, driven by the vast availability of high quality data. Now is the time to put this data to use and fuel innovative initiatives with new tools.
When our clients started saying that they had no way to properly test new ideas, or that their timetable systems produced timetables that worked on paper, but not in real life, we decided to rise to the challenge. We have dedicated ourselves to boost efficiency and innovation by using data to its full potential and NetworkSim is one of the key parts of that vision.
We combine our years of experience in the mobility sector with creativity, pure data science and our love for coding to build the applications of tomorrow.
The availability of enormous amount of computing power in combination with multiple new data sources ask for a new approach in the industry. Gone are the days trial and error in network planning.
Utilising the strength of simulation tools while addressing weaknesses in the existing network planning software options is what we're all about. Our goal is to provide insights and boost improvements while doing what we love.
Meet the team behind NetworkSim.
Head Development Europe
Head Development Asia Pacific