Study of bend-taking behaviour of motorbike riders application to training and intelligent transport systems – VIROLO++

The project is structured around the use of instrumented PTWs, the instrumentation being more or less heavy depending on the studies to be conducted. The data required for the reconstruction of “precise” trajectories and the calculation of the “risk function” are acquired using existing instrumented PTW, with the addition of systems that allow “precise” localisation. The study of the rider/PTW interactions required a much heavier instrumentation in order to measure the pressures exerted by riders on their PTWs at the various contact points. Thanks to the acquired data, the project aims at the design of: 1- a cybernetic model of steering control that is as close as possible to that achieved by PTW riders. This model will be based on current knowledge in cognitive ergonomics, behavioural neuroscience and human biomechanics. The goal to be achieved - and, indeed, the main scientific challenge to be overcome - is to propose a biologically plausible model that is accurate enough to make predictions about rider behaviour 2- the assessment of bend-taking practices through the use of “off-line” tools in pre-test and/or continuous training. This will be carried out using “on-line” recorded data, which is gathered through embedded sensors. Precise “actual” trajectories will then be compared with “safety” trajectories and help trainers in their pedagogic, 3- the development of a curriculum that delivers training/retraining in safe cornering procedures using a “low-cost” riding simulator, 4- the design of “on-line” tools for risk quantification when approaching a bend, suitable for use by novice riders, particularly during pre-test training and the immediate post-test period; 5- the refinement of the “high-end” riding simulators, thanks to new knowledge on riders' actual bend-taking practices (i.e., the interactions between riders and PTWs, for a range of the riding population).

Three PTW have been instrumented, thanks to the design of a new hardware architecture, and used to conduct experiments on road with trainees and trainers. The instrumentation used low cost and high cost redundant sensors, the aim being to compare them and to identify the minimal subset of sensors required for the reconstruction of the trajectories. A database has been completed. Using the same hardware architecture, but with several additional sensors, a PTW has been instrumented for the recording of the interactions between the rider and the motorbike. The pressures exerted by the rider on each half handlebar, on the foot-pegs but also on the tank and the saddle are collected. 3 IMUs are also used to estimate the position of the rider upper body, including the head. Several experiments have been conducted on track. A database has been completed. Experiments have been conducted to identify the dynamic parameters of the motorbikes, required for the validation of the PTW numerical modeling. A software has been designed that allows to browse the various collected data, and to export all or sub-part of the database. A software has been designed to view, using smartphones or tablets, the trajectories achieved by the riders on road or on tracks. Last, a web site has been setup (prototype).

The two database (road, track) are currently used by partners to 1) design, test, and tune an algorithm aiming at a precise reconstruction of the achieved trajectories, 2) understand the rider/vehicle interactions when taking bends, towards the design of a cybernetic model, and 3) tune PTW dynamic model and design a risk function for approaching bends, towards an ARAS dedicated to trainees and novice riders,

Several peer reviewed conference papers have been produced, peer reviewed journals are expected in the next months.

The Powered Two Wheelers (PTW) riders are the most vulnerable road users. Either in Europe or USA, 14% of the road
fatalities relates to motorcycles (IRTAD 2012). Motorcyclists involved in accidents are 20 times more likely to be killed than
car drivers. Novice riders are particularly vulnerable: young riders aged between (18-24 years) old are three times more
likely to be involved in an accident than people aged between (45-64) years old (ONISR1). Both social and economic
issues are important, the death of a young man/women costs on average 1.4 million € to the community (ONISR).
Despite this high-risk level, the market of the PTW has exploded over the last decade. The increase in the number of
motorcyclists over the last years can be explained by the benefits in terms of mobility in congested cities. Unless a great
economic turmoil, the trend should not change.
Many large-scale research programs have been undertaken in Europe and abroad to understand the factors contributing
to crashes. In particular, the MAIDS and RIDER projects allowed to characterize accidents situations, which paved the
way to other projects such as SAFERIDER (FP7) for the development of ITS. 2BESAFE (FP7) aimed to study the
motorcyclists behaviour and the behavioural and ergonomic factors contributing to motorcycle crashes. The French
ANR/Predit SUMOTORI and DAMOTO collaborative projects proposed an automatic fall detection algorithm for early
inflating of a wireless air-bag jacket. SIM2CO+ (ANR/Predit) currently aims at identifying the risky situations experienced
by novice motorcyclists who have just passed their test, in order to improve pre-test training in France.
The goal of the VIROLO++ project is to develop tools and methods suitable for the study of the riders’ behaviour and for the
understanding of the way they interact with their vehicle when negotiating a bend. Particularly, the aim is to improve
knowledge about this specific manoeuvre, because the loss of control in a bend is more than 50% of single vehicle
accidents.
The project aims to reduce the mortality (and disabling injuries) due to loss of control in bends through the development of:
1. Knowledge about how motorcyclist effectively take bends and the resulting trajectories. Experiments will objectively
measure the motorcyclist behaviour and the rider-vehicle interaction;
2. A cybernetic model of steering control that will represent as close as possible what is achieved by the rider and allow to
make suggestion for research on active safety systems operating in bends (ie devices taking all or part of the vehicle
control);
3. Offline tools for the objective evaluation of bend taking practices that should be used for the initial training and
retraining. The goal is to reconstruct "fine" trajectories from recorded data and to compare them to safe reference
trajectories, taking into account the visibility in the bend but also criteria minimization of the fuel consumption;
4. A curriculum on a low cost motorcycle simulator that can be used for training and retraining safe practices in taking a
bend (see item 3);
5. Online tools for the quantification of risk when approaching the bend and after, tools that can be used by novice riders
during and after training;
6. A improvement of high performance motorcycle simulators designed by Ifsttar thanks to the knowledge on bend negociation
practices among a diverse population of motorcyclists.

1 ONISR: French National Inter-departemental Observatory on Road Safety