The final release of Aimsun Next version 8.4, came out on May 17, 2019, and includes significant improvements to Aimsun Next traffic modeling software, with a special focus on enhancements for modeling for connected and autonomous vehicles.
The latest minor update was Aimsun 8.4.2, released in February 2020.
The next release will be Aimsun Next 20, scheduled for release in spring 2020.
This is a quick guide to the main highlights, followed by a breakdown of all enhancements and new features – you can also read this in the Aimsun Next software itself – just look up the New Features Guide after installation.
For further information please contact us at email@example.com.
New CAV modeling enhancements include:
The list of new features is not limited to CAV applications, and includes:
Further details on all of these new features below.
Autonomous vehicles are controlled differently and have different decision-making behavior from normal vehicles in the traffic network. Up until now, the physical and dynamic properties of a vehicle in an Aimsun Next microsimulation have been specified by vehicle type: the size of a vehicle and its speed and acceleration have been determined by parameters set by type. Now, the decision-making parameters are also varied by vehicle type to enable a modeler tasked with investigating the effect of different vehicle behavior on the traffic network and on the other vehicles in the network.
Where the decision-making is more complex than can be represented by parameter changes (such as simulating autonomous control software as it makes complex decisions about its path through the network and its reactions to vehicles in proximity to it), the controller can now include themselves in the simulation using a “Hardware-in-the-Loop” method to present the controller with data about adjacent vehicles and to implement the actions of the controlled vehicle in the simulation.
Connected vehicles transmit and receive more information about their activity than conventional vehicles, and this information is also available to traffic control centers though ITS infrastructure. This enables new forms of vehicle behavior through V2V communications – i.e., by platooning or by collaborative maneuvers. It also enables new levels of traffic network control facilitated by the more detailed data available from connected vehicles through V2VI communications.
In summary, the three main groups of changes made in Aimsun Next 8.4 are: first, to add to the parameters that control vehicle behavior by vehicle type; second, to implement a new, and easy-to-use, interface for external control agents; and third, to enable V2X communications in the Aimsun Next API.
The extensions to the vehicle behavior by type, take the behavior appropriate to a road section or to a turn and allow vehicles to modify that by type. For example, the Distance Zones leading up to a junction specify the point where vehicles start to make their lane choice for their maneuver at that junction and where the urgency of that change increases. These distances have previously been initially determined using default values for the road type leading into the junction and modified locally for a specific junction. Now they can be modified with a factor appropriate to the vehicle type, so an autonomous vehicle can be programmed to make its lane maneuvers in a more conservative manner (earlier than other vehicles), or in a more aggressive manner (later than other vehicles). The parameters that can be set in the Vehicle Type Editor are:
Distance Zones: The Distance Zones that control where decisions are made about which lane is required can be modified by type.
Cooperation and Aggressiveness: The Gap Acceptance Model for Lane Changing parameters that control the size of gap that a vehicle requires to make a lane change can be modified by type.
Imprudent Lane Changes: The probability of using an unsafe gap can be set by type. This option allows a vehicle to accept a gap that requires it, or its follower, to brake up to twice their maximum deceleration.
Overtake Speed Threshold and Lane Recovery Speed Threshold: These parameters control a vehicle’s desire to overtake by making a lane change on a multi-lane carriageway.
Stay in Lane: This parameter controls the tendency to move back to the slower lane after overtaking.
Safety Margin: In the Junction Give Way Model, this parameter controls how close vehicles may pass when assessing safe gaps to move into.
The External Agent Interface (EAI) is designed to introduce an externally controlled vehicle into an Aimsun Next simulation and have that vehicle guided by the actions of, for example, a human driver in a simulator, an autonomous vehicle controller, or by an experimental control system being tested in a simulation environment.
The data exchanged via the EAI is based on geographic locations expressed as XY, the traffic network expressed as lane co-ordinates rather than being based on a simulated representation of road sections and turns. This means that the external control logic does not require detailed knowledge of how Aimsun Next models the traffic network; it can continue to use its own network model. Data exchange relies solely on there being a shared common co-ordinate system. The external vehicle is positioned in the traffic network in the simulation and the other vehicles in the simulation, which are controlled by Aimsun Next as normal, will then react to the presence of this external vehicle by following it, collaborating with its lane change maneuvers, and including it in their assessment of gaps at junctions in the same way as they react to other simulation vehicles.
The EAI uses Google Protocol Buffers for data transfer, ensuring cross-platform compatibility and also allowing the controller developer to use any programming tools and languages supported by Google. Aimsun does however also provide a simple interface, single vehicle interface, using a Windows DLL, usable in a Visual Studio project or from Matlab. The data transferred from the simulation to the external agent is the location of the vehicles and pedestrians surrounding the agents and the state of nearby traffic signals. The data transferred to the simulation from the external agent is the new speed, heading and position of the controlled vehicles.
The EAI differs from the Aimsun Next API and the Aimsun Next microSDK in that it does not require the external controller to include any of the simulation artefacts such as road sections, nodes or turns; the external agent is not constrained to work within those structures. All that is required, is a common coordinate system and a common map of the traffic network.
The EAI is licensed separately from Aimsun Next Editions. Please contact firstname.lastname@example.org to acquire the EAI plug-in, the simplified external agent DLL with sample code, and the protocol buffer specification file which enables platform independent coding and use with multiple external vehicles.
If you have developed the vehicle control logic in SCANeR, or you want to drive a vehicle in the microscopic simulation, we have completely redesigned the interface with this driving simulation software (requires a separate license).
The Driving Simulation Interface is a fully supported and redesigned TCP/IP connection between SCANeR and Aimsun Next. This interface allows a choice between ‘traditional’ driving simulator applications, where a human driver sits down at a console and controls a vehicle within a microscopic simulation, or an automotive simulator, which bypasses the human element entirely and creates in SCANeR virtual sensors and an autonomy stack to take control of a vehicle.
For testing connected vehicle applications, we are releasing a technical preview of the V2X Software Development Kit (this requires a separate license).
A Vehicle Ad Hoc Network (VANet) is an ephemeral network spontaneously created by a collection of connected vehicles in proximity to each other or in proximity to a similarly connected road-side unit (RSU). The communications are generically referred to as V2V for Vehicle-to-Vehicle communications, V2I for Vehicle-to-Infrastructure communications, or, when both are operating together, V2X communications. The V2X SDK, introduced in Aimsun Next 8.4 as a technology preview, is intended to include VANet communications in an Aimsun Next simulation by exchanging messages between vehicles and roadside units where those messages are based on industry standard protocols.
The design principle behind the Aimsun Next V2X SDK is to provide an extensible system in which standardized message protocols are pre-programmed and to which new, experimental protocols may be added. Similarly, recognizing the complexity of the task, the transmission and VANet membership protocols are simulated as simple communications channels, with a range and a stochastic probability of successful transmission, while the on-board units (OBU) in the vehicles are also implemented by default as simple transmitters and receivers.
For more complex studies that focus on the communications as much as the vehicle and management actions, the modeler may replace the channels and OBUS to include more of the communications protocol details. The messages received by the road-side units and the vehicles are intended to be used as inputs to a “Rules Engine”, which will then influence the management of the traffic network or the actions of each vehicle according to the data received from other vehicles in the VANet.
The V2X SDK is licensed separately from Aimsun Next editions. Please contact email@example.com to acquire the V2X plug-in, which includes the user interface required to generate the OBUs, the RSUs and the Traffic Management Centre, as well as to link them by channel/message type. The V2X SDK also includes the API components required to implement a “Rules Engine” in the vehicle or RSU.
Please note that a deep knowledge of programming is needed to use this module.
Aimsun Next 8.4 now includes an interface to Sitraffic Office from Siemens.
Sitraffic Office is a modular software suite from Siemens AG, which provides tools for traffic management based on a shared database of network assets and traffic data. Sitraffic Office provides workflow tools for integrated junction design, route planning, and network management. Sitraffic Office also integrates tools for dynamic signal actuation and coordinated signal control across junctions.
With the release of Aimsun Next 8.4, data exported from Sitraffic Office can be imported into Aimsun Next to create a new model, including the traffic network, the signal controls, the traffic demand and public transport demand, and the pedestrian network. Subsequently, after import, when a microsimulation model is run, the signals in the model can be controlled by a Siemens PDM Signal controller through an external interface to junction control software created as a part of the Sitraffic import.
The Distribution Scenario now includes a second type of experiment that implements a Destination Choice Model; you can use this in the Distribution Step of a four-step demand model instead of the existing Gravity Distribution Model.
Destination choice models are a type of trip distribution or spatial interaction model that is formulated as a discrete choice model, typically logit models, which assign trips through probabilities by maximizing the utility of each traveler of going from each origin to each destination. The key to this model is the assumption that the travelers select their destination based on the utility it has for them, while trying to maximize this factor.
The Destination Choice Model is now an option when you create a Distribution Experiment.
The UI recognises when a HiDPI screen is in use (known as Retina Display on a Macintosh) and to present appropriately scaled icons and images.
A new Dark Mode to work with Dark Mode Windows themes.
We’ve added a Dynamic Models Tab to the Centroid Editor to configure the centroid parameters for a Park-and-Ride Traffic Action.
The previous version of Aimsun Next already included the option to record and replay a mesoscopic simulation but only at a fixed interval of 5 seconds. This interval is now configurable in the replication editor.
We’ve added three new parameters to the StaticInfVeh structure returned by the AKIVehTrackedGetStaticInf call and the AKIVehGetVehicleStaticInfSection call:
We added the following functions to modify the parameters related to the vehicle’s lane changing behavior:
We’ve added scripting functions to define all road section traffic management actions (such as speed change) as a group rather than individual road sections.