Data Sets For Enhancing Safety Of Heavy Vehicles In Challenging Conditions And Ideas Of Enhancing Friction Estimation Safety And Shortening Essay Length.

We are making the use of data analytics and learning to enhance the safety (and stability) performance of heavy vehicles under challenging road and weather conditions

work on vehicle safety for heavy vehicles under challenging road conditions. You will then have to find the data sets used in those papers or the references therein and also search for more papers/data sets on your own in addition to the links/papers that I send. Those data you find will later be used for stimulation and integrated into python and other programs:
from the results it will be used machine learning on those data sets for the final results. so your task is to come up with ideas how machine learning can be used here as well so the most important here is new idea for enhancing friction estimation safety for heavy vehicles

Speed

1.Data sets

Safety of heavy vehicles in challenging conditions is very crucial. One of the best ways of improving safety of these vehicles is by creating appropriate simulations so as to identify the right combination of parameters, variables or strategies that can be applied in enhancing safety of heavy vehicle drivers and other road users. Many studies have found various data sets that can be used to create simulation models for estimating and improving safety of heavy vehicles. Some of these data sets include the following:

Speed

Speed has a significant impact on safety of heavy vehicles. When heavy vehicles travel at very high speed, they become vulnerable to high damage in the event of an accident because the driver cannot control them in case of an emergency. Driving risks increases with increasing speed standard deviation of the vehicle. In most cases, heavy vehicles tend to reduce their speed when travelling on comparatively steep longitudinal grades so as to enable the driver have more control of the vehicle and reduce accident risks (Bassan, 2016). The heavy vehicles should travel at recommended speed especially when on steep slopes.

Geometric design elements

Safety of heavy vehicles is largely affected by various geometric design elements of the road. Some of the data sets that should be used include: horizontal curvature (transition curve, super-elevation, curve radius, sight distance, etc.), vertical alignment/grade (gradients, crest curves, sag curves and vertical curves), cross slope, etc. In general, various data about the complexity of road alignment should b considered.

Shoulder width

When the rod width is small, the driver cannot control the vehicle properly. Small road width also increases the likelihood of heavy vehicles colliding with other vehicles such as passenger cars. Therefore lanes should be of adequate width so as to improve safety of heavy vehicles.

Loading

Many studies have shown that the risk of infrastructure damage and traffic accidents increase when heavy vehicle exceed the legal and acceptable weight limits. When a heavy vehicle is overloaded, its kinetic energy increases significantly thus increasing impact forces together with damage to the infrastructure and other vehicles – in the event of an accident (Jacob & Beaumelle, 2010). Different countries have varied maximum permitted weight limits for heavy vehicles. Therefore various loadings should be tested so as to determine the suitable mass limits that the heavy vehicle should be allowed to carry without exceeding legal limits.

Vehicle characteristics

There are several vehicle characteristics that can be used to improve safety of heavy vehicles in challenging conditions (Tagesson, 2014). Some of these include: braking system, steering system, driver assistance systems, anti-roll bars, and other sensing systems, which are essential in controlling directional stability of heavy vehicles (Tagesson, et al., 2016). These systems are very crucial in situations such as when cornering or when the vehicle lose control (such as lateral stability and rollover) (Trigell, et al., 2017). Drivers should understand and use them effectively to improve safety on the road.

Geometric design elements

Tunnels

Movement of heavy vehicles through tunnels is largely restricted as drivers cannot perform some actions such as a U-turn maneuver. Therefore when a road has tunnels, any physical obstacles that may restrict maneuverability of heavy vehicles should be prevented or minimized (Caliendo, et al., 2013). The data to be used when creating road tunnels’ crash-prediction models include: tunnel length, number of lanes, annual average daily traffic per lane, presence of sidewalk, and percentage of heavy vehicles. This data can be analyzed used methods such as Bivariate Negative Binomial regression model, Random Effects Binomial regression model, Negative Multinomial regression model, Inverse Gaussian regression model, Maximum Likelihood Method, and Cumulative Residual method, among others (Meng & Qu, 2012).

Road conditions or pavement surface characteristics

The condition of the road also affects the safety of heavy vehicles. When the road is in poor condition, the risks of accident are very high. For instance, a road that is full of potholes or one that is worn out makes it difficult for the driver to control the vehicle. Poor roads also reduces grip between the vehicle tires and the road, making it easy for the vehicle to slip. For instance, when friction level is different between the right and left side of the vehicle, even braking becomes a challenge for drivers (Tagesson, et al., 2014). It is therefore important to create simulations showing high possibility of heavy vehicle accidents on poor roads as this may drive governments and other relevant private stakeholders to maintain roads.

Road lighting, light condition of the vehicle and weather conditions

When weather conditions, light condition of the vehicle and road lighting are poor, the heavy vehicle and driver cannot perform as expected. Poor weather and lighting conditions make it difficult for the driver to see clearly and may accidently hit oncoming vehicles, pass over bumps at high speed or find it difficult to negotiate a corner. Therefore road lighting, light condition of the vehicle and weather conditions should be used to create simulations that will determine when and how the driver should use lighting of the vehicle, depending on daylighting level (which may be affected by fog, excess smoke, heavy rain, etc.). Weather conditions such as strong winds and thunderstorms also affect safety of heavy vehicles and should be used to create safety models of heavy vehicles. These conditions largely affect drivers’ behavior, ability and flexibility to handle unstructured elements (Tagesson, 2017).

Road signs and control systems

Shoulder width

Properly painted and signalized roads are less risky to accidents because drivers know in advance what is ahead and what to do so as to avoid accidents. These systems are very essential along steep or meandering roads, at intersections and mid-blocks, etc. Therefore data about control systems and road signs should be used to create models that improve safety of heavy vehicles.

Safety management interventions

This is another very important, but sometimes overlooked, data set that has a huge impact on safety of heavy vehicles. When designing a comprehensive plan to improve safety of heavy vehicles in challenging conditions, it is important to consider the following factors: safety training, management commitment, journey planning and scheduling, environmental and vehicle conditions, involvement of employees, support/communication, incentives, type of freight (freights such as flammable products increases severity of damage in case of an accident), and safety technologies of the vehicle, among others (Mooren, et al., 2014).

2.Ideas of enhancing friction estimation safety

Friction is a major contributing factor to safety of heavy vehicles. Therefore it is important to calculate tire-road friction coefficient when developing safety models for heavy vehicles in challenging conditions. Some of the ways of enhancing friction estimation include:

Using measured values from yaw rate sensors and wheel angular velocity – this involves identifying longitudinal and lateral velocities using Kalman filter observer, estimating longitudinal and lateral tire forces using recursive least square algorithm, and estimating friction coefficient using estimated values from the previous stages and multilayer perception neural network.

System models – this is where friction estimation is done using models such as single wheel model, anisotropic brush model, vehicle lateral dynamics, enhanced adaptive observer, recursive least square method, etc.

Environment and road sensing – this involves use of sensors to determine various parameters affecting tire-road friction.

Cooperative methods – this involves use of information shared between vehicles and road infrastructure. The method combines environmental and road sensing methods together with vehicle dynamics observation methods.

Vehicle dynamics observation techniques – this method is used to determine tire behavior, depending on tire loading. Some of these methods include: longitudinal dynamics based, lateral dynamics based, and combination of lateral and longitudinal dynamics (Prokes, 2015).

Machine learning – this ideas involves used of data collected from weather stations and historical friction data collected from connected vehicles. The road friction coefficients are then calculated using models such as support vector machine, logistic regression and neural networks (Ghazaleh, Nasser  &  Zenuity, 2017).

In general, ideas of estimating friction for heavy vehicles entail use of in-vehicle measurements to observe tire behavior.

References:

Bassan, S., 2016. Overview of Traffic Safety Aspects and Design in Road Tunnels. IATSS Research, 40(1), pp. 35-46.

Caliendo, C., De Guglielmo, M. & Guida, M., 2013. A Crash-Prediction Model for Road Tunnels. Accident Analysis & Prevention, Volume 55, pp. 107-115.

Jacob, B. & Beaumelle, V., 2010. Improving Truck Safety: Potential of Weigh-in-Motion Technology. IATSS Research, 34(1), pp. 9-15.

Meng, Q. & Qu, X., 2012. Estimation of Rear-End Vehicle Crash Frequencies in Urban Road Tunnels. Accident Analysis & Prevention, Volume 48, pp. 254-263.

Mooren, L., Grzebieta, R., Williamson, A. & Friswell, R., 2014. Safety Management for Heavy Vehicle Transport: A Review of the Literature. Safety Science, Volume 62, pp. 79-89.

Ghazaleh, P.G., Nasser, M., Zenuity, A.B., 2017. Road Friction Estimation for Connected Vehicles Using Supervised Machine Learning. Goteborg, Sweden: Chalmers University of Technology.

Prokes, J., 2015. Realtime Estimation of Tyre-road Friction for Vehicle State Estimator. Goteborg, Sweden: Chalmers University of Technology.

Tagesson, K., 2014. Truck Steering System and Driver Interaction. Goteborg, Sweden: Chalmers University of Technology.

Tagesson, K., 2017. Driver-centered Motion Control of Heavy Vehicles. Goteborg, Sweden: Chalmers University of Technology.

T Tagesson, K; Eriksson, B; Hulten, J; Pohl, J; Laine, L; Jacobson, B., 2016. Improving Directional Stability Control in a Heavy Truck by Combining Braking and Steering Action. Goteborg, Sweden, Chalmers University of Technology.

Tagesson, K., Jacobson, B. & Laine, L., 2014. Driver Response to Automatic Braking under Split Friction Conditions. Tokyo, JSAE.

Trigell, A., Rothhamel, M., Pauwelussen, J. & Kural, K., 2017. Advanced Vehicle Dynamics of Heavy Trucks with the Perspective of Road Safety. International Journal of Vehicle Mechanics and Mobility, 55(10), pp. 1572-1617.