Products Consultancy
Home Products OmniTRANS Planned developments
Print

Research and Development


Research and development is a core activity of OmniTRANS.  Every year, our team of researchers comes up with new ideas and innovation for improving transport modelling and facilitating its use within the urban planning and management context. Our efforts focus on two elements: state-of-the-art transport modelling and productivity enhancement. Our vision is to bring traffic modelling within the reach of all urban and transport planners for the benefits of the whole community. Our constantly evolving product is the realisation of this vision.

 

State-of-the-Art Transport Modelling

 

Productivity Enhancement

 

The Way Forward

The results of these research and development efforts find their way into new versions of OmniTRANS.  Here is an overview of the main features that are planned for future releases  versions of OmniTRANS.

 

Version 6.0

OmniTRANS 6.0 will introduce the following new features: 

  • Multi-core Processing: The static assignment model, OtTraffic, and the public transport assignment model, OtTransit, support the use of multi-core processing. A performance improvement factor varying between 2 and 3 can be expected using a quad core processor.  
  • Cordon Assignment: The static assignment model, OtTraffic, allows you to obtain a cordon matrix associated with a pre-defined cordon. Afterwards, you can directly carry out an assignment using the cordon matrix and corresponding subnet allowing you to quickly focus on a particular area of interest. The interface has been modified to facilitate the process. The polygon associated with a selection can be saved, making it possible to display cordons alongside other objects in the network graphics. 
  • User-defined Cost Function: OtTraffic has a new property “costFunction” allowing the user to implement its own cost function using a Ruby method or class.  
  • User-defined Convergence Function : OtTraffic has a new property “convergenceFunction” allowing the user to implement its own convergence function using a Ruby method or class. When defined, OtTraffic uses the convergence function to control the criteria at which OtTraffic will stop during its iterative process.  
  • Capacity Restraint for Public Transport: The new OtTransit can account for the effect of crowding on transit lines when carrying out its assignment.  It is possible to define a crush capacity factor for each transit lines to control the perceived discomfort of passengers as the service gets crowded. The algorithm calculates a higher experienced travel time on highly saturated transit lines. This calculation is dependent on expected occupation, the number of seats available, the crush capacity of the service, the frequency and reliability of service on the line.  
  • Multiple Egress Routes: In addition to multiple access candidates, the new OtTransit considers multiple egress candidates on the same transit line when choosing suitable multi-modal travel routes between an origin and a destination.
  • Area and Link Geographical Contours: The methodology for managing, storing and retrieving the geographical contours associated with links and areas has been re-engineered. The gains are a faster loading of complex transport networks and faster import and export operations. 
  • Improved network management operations: The methodology for creating, editing and deleting network objects has been redesigned and harmonised. A transport link with its detailed geographical contour can be quickly added using successive mouse clicks. 
  • GUI Rendering of Transit Lines:  Transit lines can be drawn without offset. Alternatively, transit lines can be grouped by sub-modes and the offset drawn by sub-mode. 
  • Projection systems: The GIS projection mechanisms have been re-engineered opening the door to several projection systems including Google’s projection system. The Views and the ability to interact with Google Maps has been improved. 
  • Zoom Level: Version 6.0 provides the means to adjust the zoom level at which a particular object type is drawn on the network display. As the user zooms in and out, different object types are displayed providing a faster network drawing response time and a rendering that is adapted to the zoom level perspective. 
  • Categorised Designs: Your lists of design and report specifications, selections and views can be organized into folders in a nested fashion greatly easing the management of large projects.   
  • Design Manager: The management of selections, views and queries have been harmonized and integrated within the Design Manager.  
  • Selection Editing: The polygon used in defining a selection can now be saved and displayed on the network view. As a result, it is possible to re-use the polygon for defining other object selections.  
  • Ruby Version 1.9: This version of Ruby brings several enhancements including faster execution. 
  • New job engine: The job engine dll has been replaced by an executable bringing improved memory management. 
  • Improved Performance: Several modifications were made under the hood bringing overall performance improvements. 
  • New and redesigned dynamic assignment class: Up to four times faster dynamic assignment algorithm.
  • Route Storage: Ability to store the assigned routes within both static and dynamic assignment. 
  • Multi-language support: Introduced in Version 5.1.6 for the Italian language, the ability to support multi-language has been formalised.

 

Static Assignment Modelling

The OmniTRANS static assignment model, OtTraffic, offers simplicity of use as well as a broad range of functionality, including detailed junction modelling. During 2009 , the model has been completely re-engineered and is now offering integrated multi-class assignment within Version 5.1. Future developments include route storage and assignment of multiple cordons.

 

Dynamic Assignment Modelling

OmniTRANS macroscopic dynamic assignment model, called MaDAM, is based on an adaptation of the fluid transmission model of Messmer and Papageorgiou employing the single-regime speed-flow-density relationships of Van Aerde.

In the last three years, Omnitrans International has invested heavily in the development of a completely new dynamic assignment framework and algorithm, called Streamline, which considerably refines the above approach. The new Streamline model offers:

  • considerable performance improvements
  • routes, route generation and route choice models
  • storage of routes for further reference outside the model
  • single routes, turn proportions and a new way of combining single routes and turn proportion based approaches
  • ability to model dynamic traffic management measures via controls
  • improved interactions with junction modelling
  • support of multi-core CPUs resulting in even faster simulation
  • a new innovative approach with an open software architecture enabling third-parties to develop their own modules

 

The new Streamline algorithm and associated functionality will appear in Version 5.2 and 5.3. 

 

Advanced Junction Modelling

OmniTRANS offers extensive junction modelling capability that is well integrated with all its assignment models: static, dynamic and public transport. Development efforts focus on easing the modelling of complex junctions, from multiple arms roundabouts to complex turbo roundabouts.

 

Schedule-based Public Transport Modelling

OmniTRANS has an innovative approach to modelling public transport based on mode chain, integrated multi-modal networks and fares. The access and egress legs as well as transfer between transit lines can be modelled, closely reflecting all of the decisions travellers make for complete door-to-door journeys. The model is suitable for modelling complex schemes such as Park & Ride, Bike & Ride and Kiss & Ride. Current development efforts consist in integrating schedule-based transport system in addition to the existing frequency based systems.

 

Disaggregated Demand Models

OmniTRANS already offers a class for discrete choice modelling which supports multinomial (MNL) and nested logit (NL) forms. Current development efforts focus on easing the use of cross-nested and incremental models within OmniTRANS.  

 

Multi core CPUMulti-core Processing

Computers with multi-core processors are now common. Omnitrans International is reviewing its core algorithms and investing in their parallelisation to take advantage of the availability of multi-core processors. In Version 5.2, the static assignment model, OtTraffic, will be able to use multi-core processors. Initial tests show some 300% performance improvements using a quad core processor.

 

Controls for Modelling Dynamic Traffic Management Measures

The upcoming new Dynamic Assignment Model Streamline will greatly facilitate the modelling of dynamic traffic management measures such as: peak lanes, speed restrictions, lane and overtaking restrictions, ramp metering, road closures, tolls, DRIP, etc. The OmniTRANS GUI will provide a new network object type, called a control, to allow the user to set and adjust those measures before the execution of a dynamic assignment.   

 

Route Storage and Route Visualisation

The upcoming Dynamic Assignment Model Streamline will be delivered with a new route generation and route choice models including a facility to store those routes. Route storage will also be part of the static assignment in Version 5.2. For large models, there can be thousands if not millions of routes. Considerable effort is being spent in developing an effective route storage mechanism that can be well integrated with the OmniTRANS PMTURI dimension concept. Innovative interfaces are under development for retrieving the routes determined by both the static and dynamic assignment and for visualising them in the network view.

 

OpenGIS Support

Interoperability with the OpenGIS standard is under development to facilitate the use of these products within the OmniTRANS application.

 

Multi-language Support

Multi-language support for a specialised application such as OmniTRANS is a challenge and expensive operation. OmniTRANS with its rich classes provides a Ruby based development language for modelling advanced transport systems. This part of the system will continue to evolve in the English language. For the user interface and messages sent to the user, mechanisms are being developed to ensure that these services can be made available in other languages if desired.