Demonstration of reduced airport congestion through pushback rate control

Airport surface congestion results in significant increases in taxi times, fuel burn and emissions at major airports. This paper describes the field tests of a congestion control strategy at Boston Logan International Airport. The approach determines a suggested rate to meter pushbacks from the gate, in order to prevent the airport surface from entering congested states and to reduce the time that flights spend with engines on while taxiing to the runway. The field trials demonstrated that significant benefits were achievable through such a strategy: during eight four-hour tests conducted during August and September 2010, fuel use was reduced by an estimated 12,250–14,500 kg (4000–4700 US gallons), while aircraft gate pushback times were increased by an average of only 4.4 min for the 247 flights that were held at the gate.     

Dynamic travel time prediction using data clustering and genetic programming

The current state-of-practice for predicting travel times assumes that the speeds along the various roadway segments remain constant over the duration of the trip. This approach produces large prediction errors, especially when the segment speeds vary temporally. In this paper, we develop a data clustering and genetic programming approach for modeling and predicting the expected, lower, and upper bounds of dynamic travel times along freeways. The models obtained from the genetic programming approach are algebraic expressions that provide insights into the spatiotemporal interactions. The use of an algebraic equation also means that the approach is computationally efficient and suitable for real-time applications. Our algorithm is tested on a 37-mile freeway section encompassing several bottlenecks. The prediction error is demonstrated to be significantly lower than that produced by the instantaneous algorithm and the historical average averaged over seven weekdays (p-value <0.0001). Specifically, the proposed algorithm achieves more than a 25% and 76% reduction in the prediction error over the instantaneous and historical average, respectively on congested days. When bagging is used in addition to the genetic programming, the results show that the mean width of the travel time interval is less than 5 min for the 60–80 min trip.     

Characterization and prediction of air traffic delays

This paper presents a new class of models for predicting air traffic delays. The proposed models consider both temporal and spatial (that is, network) delay states as explanatory variables, and use Random Forest algorithms to predict departure delays 2–24 h in the future. In addition to local delay variables that describe the arrival or departure delay states of the most influential airports and links (origin–destination pairs) in the network, new network delay variables that characterize the global delay state of the entire National Airspace System at the time of prediction are proposed. The paper analyzes the performance of the proposed prediction models in both classifying delays as above or below a certain threshold, as well as predicting delay values. The models are trained and validated on operational data from 2007 and 2008, and are evaluated using the 100 most-delayed links in the system. The results show that for a 2-h forecast horizon, the average test error over these 100 links is 19% when classifying delays as above or below 60 min. Similarly, the average over these 100 links of the median test error is found to be 21 min when predicting departure delays for a 2-h forecast horizon. The effects of changes in the classification threshold and forecast horizon on prediction performance are studied.     

Tarmac delay policies: A passenger-centric analysis

In this paper, we analyze the effectiveness of the 2010 Tarmac Delay Rule from a passenger-centric point of view. The Tarmac Delay Rule stipulates that aircraft lift-off, or an opportunity for passengers to deplane, must occur no later than 3 h after the cabin door closure at the gate of the departure airport; and that an opportunity for passengers to deplane must occur no later than 3 h after the touchdown at the arrival airport. The Tarmac Delay Rule aims to protect enplaned passengers on commercial aircraft from excessively long delays on the tarmac upon taxi-out or taxi-in, and monetarily penalizes airlines that violate the stipulated 3-h tarmac time limit. Comparing the actual flight schedule and delay data after the Tarmac Delay Rule was in effect with that before, we find that the Rule has been highly effective in reducing the frequency of occurrence of long tarmac times. However, another significant effect of the rule has been the rise in flight cancellation rates. Cancellations result in passengers requiring rebooking, and often lead to extensive delay in reaching their final destinations. Using an algorithm to estimate passenger delay, we quantify delays to passengers in 2007, before the Tarmac Delay Rule was enacted, and compare these delays to those estimated for hypothetical scenarios with the Tarmac Delay Rule in effect for that same year. Our delay estimates are calculated using U.S. Department of Transportation data from 2007. Through our results and several sensitivity analyses, we show that the overall impact of the current Tarmac Delay Rule is a significant increase in passenger delays, especially for passengers scheduled to travel on the flights which are at risk of long tarmac delays. We evaluate the impacts on passengers of a number of rule variations, including changes to the maximum time on the tarmac, and variations in that maximum by time-of-day. Through extensive scenario analyses, we conclude that a better balance between the conflicting objectives of reducing the frequency of long tarmac times and reducing total passenger delays can be achieved through a modified version of the existing rule. This modified version involves increasing the tarmac time limit to 3.5 h and only applying the rule to flights with planned departure times before 5pm. Finally, in order to implement the Rule more effectively, we suggest the tarmac time limit to be defined in terms of the time when the aircraft begin returning to the gate instead of being defined in terms of the time when passengers are allowed to deplane.     

Fueling for contingencies: The hidden cost of unpredictability in the air transportation system

The aviation community is increasing its attention on the concept of predictability when conducting aviation service quality assessments. Reduced fuel consumption and the related cost is one of the various benefits that could be achieved through improved flight predictability. A lack of predictability may cause airline dispatchers to load more fuel onto aircraft before they depart; the flights would then in turn consume extra fuel just to carry excess fuel loaded. In this study, we employ a large dataset with flight-level fuel loading and consumption information from a major US airline. With these data, we estimate the relationship between the amount of loaded fuel and flight predictability performance using a statistical model. The impact of loaded fuel is translated into fuel consumption and, ultimately, fuel cost and environmental impact for US domestic operations. We find that a one-minute increase in the standard deviation of airborne time leads to a 0.88 min increase in loaded contingency fuel and 1.66 min in loaded contingency and alternate fuel. If there were no unpredictability in the aviation system, captured in our model by eliminating standard deviation in flight time, the reduction in the loaded fuel would between 6.12 and 11.28 min per flight. Given a range of fuel prices, this ultimately would translate into cost savings for US domestic airlines on the order of $120–$452 million per year.     

Fast computing and approximate fuel consumption modeling for Internal Combustion Engine passenger cars

This article presents a fuel consumption model, SEFUM (Semi Empirical Fuel Use Modeling), and its comparison with three models from the literature on a 600 km experimental database. This model is easy to calibrate with only a few required parameters that are provided by car manufacturers. The test database has been built from 21 drivers who drove in two conditions (normal and ecodriving) on a 15 km trip. For the model evaluation, three indicators have been selected: instantaneous fuel use root mean square error, cumulated error and computation time in order to evaluate the accuracy both in cumulated and instantaneous fuel use and to estimate computation time of each model. Results tend to prove that the model is able to compute rapidly (maximum of 1500 simulated kilometers under Matlab) in comparison to all other models while ensuring a high accuracy and precision for cumulated and instantaneous fuel use.     

Characteristics of air pollutants at near and far field regions of a national highway located at an industrial complex

This paper presents the characterization of air quality monitored at near field region (NFR) and far field region (FFR) of a national highway located at an industrial complex. The pollutants such as PM₁₀, SO₂ and NO₂ were monitored in two campaigns (11th September to 18th October 2012 and 18th January to 17th February 2013). The 24 h average PM₁₀ concentration at NFR and FFR were found to be 86.69 ± 18.56 μg/m³; 73.16 ± 16.21 μg/m³ and 89.44 ± 18.69 μg/m³; 81.91 ± 16.42 μg/m³, respectively during first and second campaign. In both the campaigns PM₁₀, SO₂ and NO₂ concentration at NFR was higher than FFR. The chemical characterization of PM₁₀ at NFR and FFR indicated the abundance of major elements such as Na (NFR = 30% and FFR = 32%), Ca (NFR = 12% and FFR = 14%) and ions namely NO₃⁻ (NFR = 71% and FFR = 68%) and NH₃⁺ (NFR = 15% and FFR = 19%). Further, at FFR, SO₄²⁻ and NO₃⁻ were found to be 18% and 35% higher than NFR indicating the conversions of SO₂ and NO₂ concentration into secondary particles. The measured SO₂ and NO₂ concentrations were 23 and 21% lower at FFR when compared to NFR confirms the secondary formation.The CALPUFF, EPA regulatory model was set up to understand the dynamics of air pollutants at the industrial complex. The predicted PM₁₀, SO₂ and NO₂ concentrations at NFR and FFR were found to be 32.31 ± 1.56 μg/m³ and 31.35 ± 1.27 μg/m³; 0.37 ± 0.21 μg/m³ and 0.06 ± 0.04 μg/m³; 12.83 ± 6.55 μg/m³ and 4.67 ± 2.77 μg/m³, respectively. The model showed moderate predictions for PM₁₀ (R² = 0.44–0.52), SO₂ (R² = 0.41–0.51) and NO₂ (R² = 0.45–0.61) concentrations.     

Using nonparametrics to specify a model to measure the value of travel time

Using a range of nonparametric methods, the paper examines the specification of a model to evaluate the willingness-to-pay (WTP) for travel time changes from binomial choice data from a simple time–cost trading experiment. The analysis favours a model with random WTP as the only source of randomness over a model with fixed WTP which is linear in time and cost and has an additive random error term. Results further indicate that the distribution of log WTP can be described as a sum of a linear index fixing the location of the log WTP distribution and an independent random variable representing unobserved heterogeneity. This formulation is useful for parametric modelling. The index indicates that the WTP varies systematically with income and other individual characteristics. The WTP varies also with the time difference presented in the experiment which is in contradiction of standard utility theory.     

Thermodynamic,environmental and economic effects of diesel and biodiesel fuels on exhaust emissions and nano-particles of a diesel engine

In this study, diesel (JIS#2) and various biodiesel fuels (BDF20, BDF50, BDF100) are used to operate the diesel engine at 100 Nm, 200 Nm and full load; while the engine speed is 1800 rpm. The system is experimentally studied, and the energy, exergy, sustainability, thermoeconomic and exergoeconomic analyses are performed to the system. The Engine Exhaust Particle Sizer is used to measure the size distribution of engine exhaust particle emissions. Also, the data of the exhaust emissions, soot, particle numbers, fuel consumptions, etc. are measured. It is found that (i) most of the exhaust emissions (except NO_x) are directly proportional to the engine load, (ii) maximum CO₂ and NO_x emissions rates are generally determined for the BDF100 biodiesel fuel; while the minimum ones are calculated for the JIS#2 diesel fuel. On the other hand, the maximum CO and HC emissions rates are generally computed for the JIS#2 diesel fuel; while the minimum ones are found for the BDF100 biodiesel fuel, (iii) fuel consumptions from maximum to minimum are BDF100 > BDF50 > BDF20 > JIS#2 at all of the engine loads, (iv) particle concentration of the JIS#2 diesel fuel is higher than the biodiesel fuels, (v) soot concentrations of the JIS#2, BDF20 and BDF50 fuels are directly proportional to the engine load; while the BDF100 is inversely proportional, (vi) system has better energy and exergy efficiency when the engine is operated with the biodiesel fuels (vii) sustainability of the fuels are BDF100 > BDF50 > BDF20 > JIS#2, (viii) thermoeconomic and exergoeconomic parameters rates from maximum to minimum are JIS#2 > BDF20 > BDF50 > BDF100.     

Carnivore incidents with U.S. civil aircraft

Wildlife incidents with aircraft are of concern in the United States as they pose a risk to human safety and economic losses for the aviation industry. Most previous research on wildlife-aircraft incidents has emphasized birds, bats, and ungulates. We queried the Federal Aviation Administration’s National Wildlife Strike Database from 1990 to 2012 to characterize carnivore incidents with U.S. civil aircraft. We found 1016 carnivore incidents with aircraft representing at least 16 species, with coyotes (n = 404) being the species most frequently struck. California and Texas had the most reported incidents and incidents were most likely to occur at night from August to November. Overall estimated damage to aircraft was US$ 7 million. Coinciding with the increase in air traffic, the rate of carnivore-aircraft incidents increased 13.1% annually from 1990 to 2012 whereas the rate of damaging incidents remained fairly constant. Due to the increase in carnivore-aircraft incidents from 1990 to 2012, we recommend further research on techniques to increase detection of carnivores and implementation and scheduled maintenance of perimeter high fences for exclusion. Additionally, we recommend increasing patrol of runways, especially during peak incident periods (July–November) and at night (2000–0600 h).     

Isocyanic acid and ammonia in vehicle emissions

Vehicles are considered to be an important source of ammonia (NH₃) and isocyanic acid (HNCO). HNCO and NH₃ have been shown to be toxic compounds. Moreover, NH₃ is also a precursor in the formation of atmospheric secondary aerosols. For that reason, real-time vehicular emissions from a series of Euro 5 and Euro 6 light-duty vehicles, including spark ignition (gasoline and flex-fuel), compression ignition (diesel) and a plug-in electric hybrid, were investigated at 23 and −7 °C over the new World harmonized Light-duty vehicle Test Cycle (WLTC) in the Vehicle Emission Laboratory at the European Commission Joint Research Centre Ispra, Italy. The median HNCO emissions obtained for the studied fleet over the WLTC were 1.4 mg km⁻¹ at 23 °C and 6 mg km⁻¹ at −7 °C. The fleet median NH₃ emission factors were 10 mg km⁻¹ and 21 mg km⁻¹ at 23 and −7 °C, respectively. The obtained results show that even though three-way catalyst (TWC), selective catalytic reduction (SCR), and NOx storage catalyst (NSC) are effective systems to reduce NOx vehicular emissions, they also lead to considerable emissions of the byproducts NH₃ and/or HNCO. It is also shown that diesel light-duty vehicles equipped with SCR can present NH₃ emission factors as high as gasoline light-duty vehicles at both, 23 and −7 °C over the WLTC. Therefore, with the introduction in the market of this DeNOx technology, vehicular NH₃ emissions will increase further.     

Capitalization of BRT network expansions effects into prices of non-expansion areas

A before and after hedonic model is used to determine the property value impacts on properties already served by the transit system caused by extensions to Bogotá’s bus rapid transit system. Asking prices of residential properties belonging to an intervention area (N = 1407 before, 1570 after) or a control area (N = 267 before, 732 after) and offered for sale between 2001 and 2006 are used to determine capitalization of the enhanced regional access provided by the extension. Properties offered during the year the extension was inaugurated and in subsequent years have asking prices that are between 13% and 14% higher than prices for properties in the control area, after adjusting for structural, neighborhood and regional accessibility characteristics of each property. Furthermore, the appreciation is similar for properties within 500 m and properties between 500 m and 1 km of the BRT.     

How wide should be the adjacent area to an urban motorway to prevent potential health impacts from traffic emissions?

In recent years, several studies show that people who live, work or attend school near the main roadways have an increased incidence and severity of health problems that may be related with traffic emissions of air pollutants. The concentrations of near-road atmospheric pollutants vary depending on traffic patterns, environmental conditions, topography and the presence of roadside structures. In this study, the vertical and horizontal variation of nitrogen dioxide (NO₂) and benzene (C₆H₆) concentration along a major city ring motorway were analysed. The main goal of this study is to try to establish a distance from this urban motorway considered “safe” concerning the air pollutants human heath limit values and to study the influence of the different forcing factors of the near road air pollutants transport and dispersion. Statistic significant differences (p = 0.001, Kruskal–Wallis test) were observed between sub-domains for NO₂ representing different conditions of traffic emission and pollutants dispersion, but not for C₆H₆ (p = 0.335). Results also suggest significant lower concentrations recorded at 100 m away from roadway than at the roadside for all campaigns (p < 0.016 (NO₂) and p < 0.036 (C₆H₆), Mann–Whitney test). In order to have a “safe” life in homes located near motorways, the outdoor concentrations of NO₂ must not exceed 44–60.0 μg m⁻³ and C₆H₆ must not exceed 1.4–3.3 μg m⁻³. However, at 100 m away from roadway, 81.8% of NO₂ receptors exceed the annual limit value of human health protection (40 μg m⁻³) and at the roadside this value goes up to 95.5%. These findings suggest that the safe distance to an urban motorway roadside should be more at least 100 m. This distance should be further studied before being used as a reference to develop articulated urban mobility and planning policies.     

Influence of driving patterns on vehicle emissions: A case study for Latin American cities

On-board real-time emission experiments were conducted on 78 light-duty vehicles in Bogota. Direct emissions of carbon monoxide (CO), carbon dioxide (CO₂), nitrogen oxides (NO_x) and hydrocarbons (HC) were measured. The relationship between such emissions and vehicle specific power (VSP) was established. The experimental matrix included both gasoline-powered and retrofit dual fuel (gasoline–natural gas) vehicles. The results confirm that VSP is an appropriate metric to obtain correlations between driving patterns and air pollutant emissions. Ninety-five percent of the time vehicles in Bogota operate in a VSP between −15.2 and 17.7 kW ton⁻¹, and 50% of the time they operate between −2.9 and 1.2 kW ton⁻¹, representing low engine-load and near-idling conditions, respectively. When engines are subjected to higher loads, pollutant emissions increase significantly. This demonstrates the relevance of reviewing smog check programs and command-and-control measures in Latin America, which are widely based on static (i.e., idling) emissions testing. The effect of different driving patterns on the city’s emissions inventory was determined using VSP and numerical simulations. For example, improving vehicle flow and reducing sudden and frequent accelerations could curb annual emissions in Bogota by up to 12% for CO₂, 13% for CO and HC, and 24% for NO_x. This also represents possible fuel consumption savings of between 35 and 85 million gallons per year and total potential economic benefits of up to 1400 million dollars per year.     

Simulation based population synthesis

Microsimulation of urban systems evolution requires synthetic population as a key input. Currently, the focus is on treating synthesis as a fitting problem and thus various techniques have been developed, including Iterative Proportional Fitting (IPF) and Combinatorial Optimization based techniques. The key shortcomings of these procedures include: (a) fitting of one contingency table, while there may be other solutions matching the available data (b) due to cloning rather than true synthesis of the population, losing the heterogeneity that may not have been captured in the microdata (c) over reliance on the accuracy of the data to determine the cloning weights (d) poor scalability with respect to the increase in number of attributes of the synthesized agents. In order to overcome these shortcomings, we propose a Markov Chain Monte Carlo (MCMC) simulation based approach. Partial views of the joint distribution of agent’s attributes that are available from various data sources can be used to simulate draws from the original distribution. The real population from Swiss census is used to compare the performance of simulation based synthesis with the standard IPF. The standard root mean square error statistics indicated that even the worst case simulation based synthesis (SRMSE = 0.35) outperformed the best case IPF synthesis (SRMSE = 0.64). We also used this methodology to generate the synthetic population for Brussels, Belgium where the data availability was highly limited.     

Introducing specific power to bicycles and motorcycles: Application to electric mobility

Electric bicycles and motorcycles have emerged as a possible way of improving the transportation system sustainability. This work’s aim was to quantify the energy consumption, the trip travel and the driving dynamics on specific routes in Lisbon, Portugal. Six electric and conventional bicycles and motorcycles were monitored, and a methodology to quantify the power required in each driving second was developed: Motorcycle and Bicycle Specific Power (MSP and BSP respectively). MSP and BSP allows characterizing energy consumption rates based on on-road data and to define real-world operation patterns (driving power distribution), as well as to benchmark the different propulsion technologies under the same baseline of specific power. For negative MSP and BSP modes, the conventional and the electric motorcycles and bicycles demonstrated a similar pattern. However, their behavior was different for positive modes, since electric technologies allow reaching higher power conditions. The methodology developed estimates accurately the energy consumption (average deviation of −0.19 ± 6.76% for motorcycles and of 1.41 ± 8.91% for bicycles). The MSP and BSP methodologies were tested in 2 Lisbon routes. For the electric motorcycle an increase in trip time (+36%) was observed when compared to the conventional one, while for the electric bicycle a 9.5% decrease was verified when compared to the conventional one. The Tank-to-Wheel (TTW) energy consumption for motorcycles was reduced by 61% when shifting to electric mobility, while a 30% Well-to-Wheel (WTW) reduction is obtained. For the electric bicycles, an additional energy use is quantified due to the battery electricity consumption.     

Wireless charging in California: Range,recharge, and vehicle electrification

This research evaluated the potential for wireless dynamic charging (charging while moving) to address range and recharge issues of modern electric vehicles by considering travel to regional destinations in California. A 200-mile electric vehicle with a real range of 160 miles plus 40 miles reserve was assumed to be used by consumers in concert with static and dynamic charging as a strict substitute for gasoline vehicle travel. Different combinations of wireless charging power (20–120 kW) and vehicle range (100–300 miles) were evaluated. One of the results highlighted in the research indicated that travel between popular destinations could be accomplished with a 200-mile EV and a 40 kW dynamic wireless charging system at a cost of about $2.5 billion. System cost for a 200-mile EV could be reduced to less than $1 billion if wireless vehicle charging power levels were increased to 100 kW or greater. For vehicles consuming 138 kWh of dynamic energy per year on a 40 kW dynamic system, the capital cost of $2.5 billion plus yearly energy costs could be recouped over a 20-year period at an average cost to each vehicle owner of $512 per year at a volume of 300,000 vehicles or $168 per year at a volume of 1,000,000 vehicles. Cost comparisons of dynamic charging, increased battery capacity, and gasoline refueling were presented. Dynamic charging, coupled with strategic wayside static charging, was shown to be more cost effective to the consumer over a 10-year period than gasoline refueling at $2.50 or $4.00 per gallon. Notably, even at very low battery prices of $100 per kWh, the research showed that dynamic charging can be a more cost effective approach to extending range than increasing battery capacity.     

Drivers’ parking location choice under uncertain parking availability and search times: A stated preference experiment

To assess parking pricing policies and parking information and reservation systems, it is essential to understand how drivers choose their parking location. A key aspect is how drivers’ behave towards uncertainties towards associated search times and finding a vacant parking spot. This study presents the results from a stated preference experiment on the choice behaviour of drivers, in light of these uncertainties. The attribute set was selected based on a literature review, and appended with the probabilities of finding a vacant parking spot upon arrival and after 8 min (and initially also after 4 min, but later dropped to reduce the survey complexity). Efficient Designs were used to create the survey design, where two rounds of pilot studies were conducted to estimate prior coefficients. Data was successfully collected from 397 respondents. Various random utility maximisation (RUM) choice models were estimated, including multinomial logit, nested logit, and mixed logit, as well as models accounting for panel effects. These model analyses show how drivers appear to accept spending time on searching for a vacant parking spot, where parking availability after 8 min ranks second most important factor in determining drivers’ parking decisions, whilst parking availability upon arrival ranks fourth. Furthermore, the inclusion of heterogeneity in preferences and inter-driver differences is found to increase the predictive power of the parking location choice model. The study concludes with an outlook of how these insights into drivers’ parking behaviour can be incorporated into traffic assignment models and used to support parking systems.     

General statistics of diesel engines’ idle time: Shunting locomotives in industrial sidings in Poland 2009…2013

Shunting locomotive/switcher (AmE)/utilization profiles are analyzed in this paper, in particular on the basis of idle time data collected in nineteen Polish industrial sidings and yards. 40 years old, diesel-electric locomotives are observed during 1000 h. Idle times related to work cycles are analyzed statistically. The percentage of the shunting locomotive daily operating time that the engine is operating at idle amounts to 70% (from 55% to 90%), and average daily idle fuels consumption amounts to 150 l a day (from 90 to 240 l a day).Many European and Asian countries still operate a significant number of similar (ChME3, e.g. S200) old, diesel-electric locomotives (almost 8000 locomotives have been produced), for moving trains over long distances and as shunting locomotives.Observed frequent short idle time periods suggest necessity of widening future scope of idling times’ analysis. Adaptation of Polish rolling stock will be possible using prior general public education about dangerous carcinogens in diesel exhaust smoke and fuel waste related to diesel engines’ idling. Simple simultaneous depiction of diesel engine power time series together with idle time could be used for educational visualization of idling among a wider audience. In the future classical aggregated idle time statistics should be supplemented by models that are more related to the variability of shunting locomotives diesel generator’s power time series, e.g. distribution of frequent short individual idle time cases.     

Fit-for-duty test for estimation of drivers’ sleepiness level: Eye movements improve the sleep/wake predictor

Driver sleepiness contributes to a considerable proportion of road accidents, and a fit-for-duty test able to measure a driver’s sleepiness level might improve traffic safety. The aim of this study was to develop a fit-for-duty test based on eye movement measurements and on the sleep/wake predictor model (SWP, which predicts the sleepiness level) and evaluate the ability to predict severe sleepiness during real road driving. Twenty-four drivers participated in an experimental study which took place partly in the laboratory, where the fit-for-duty data were acquired, and partly on the road, where the drivers sleepiness was assessed. A series of four measurements were conducted over a 24-h period during different stages of sleepiness. Two separate analyses were performed; a variance analysis and a feature selection followed by classification analysis. In the first analysis it was found that the SWP and several eye movement features involving anti-saccades, pro-saccades, smooth pursuit, pupillometry and fixation stability varied significantly with different stages of sleep deprivation. In the second analysis, a feature set was determined based on floating forward selection. The correlation coefficient between a linear combination of the acquired features and subjective sleepiness (Karolinska sleepiness scale, KSS) was found to be R = 0.73 and the correct classification rate of drivers who reached high levels of sleepiness (KSS ⩾ 8) in the subsequent driving session was 82.4% (sensitivity = 80.0%, specificity = 84.2% and AUC = 0.86). Future improvements of a fit-for-duty test should focus on how to account for individual differences and situational/contextual factors in the test, and whether it is possible to maintain high sensitive/specificity with a shorter test that can be used in a real-life environment, e.g. on professional drivers.