The inefficiency of Brazil’s logistics system to handle the outflow of farm products impairs the sector’s performance. From the economic perspective, the high freight rates put the country’s agricultural sector at a competitive disadvantage in the global market. Besides this, the reliance on highway transport causes problems of an environmental (high ${\mathrm{CO}}_2$ emissions) and social nature (large occurrence of accidents). As an alternative to this dependence, Brazil has an extensive network of actually and potentially navigable rivers that could serve as a key element to improve the performance of the agricultural sector. Therefore, the objective of this study is to analyze the possible influence of the Tapajós-Teles Pires Waterway on the sustainable performance of 19 Brazilian soybean export routes, using data envelopment analysis (DEA). Of these, only three would use the Tapajós-Teles Pires Waterway, which is not yet in operation. The origins of the routes are four soybean-producing centers located in Mato Grosso, the Brazilian state that produces the most soybeans, destined for 12 seaports. The inputs considered are Accident Cost and CO2 Emissions, while the output is Cargo Carried per US$. For this, we use the DEA-VRS (Variable Return Scale) input-oriented model. The results obtained from the DEA model revealed satisfactory performance of the routes using the Tapajós-Teles Pires Waterway from three of the four production centers. In Campo Novo do Parecis, some 16% (471 km) of the route is composed of highways and the rest of the waterway (84%, 1576 km). In Sorriso, only 3% (85 km) of the route consists of highways, while waterway transport is responsible for 97% (2388 km), and in Querência, 24.55% of the route to the Port of Barcarena consists of highways. The use of these routes and the combination of the two transport modes presented the best efficiency indexes, making development of the waterway a key element to improve the outflow of soybeans for export. The results of this study may help Brazilian authorities to define public policies and apply resources to improve the deficiencies of the transportation system, which can lead to a better balance of the Brazilian transportation matrix.

Road transport is the principal means of transporting freight and passengers in most developing countries, but several factors, both alone and in conjunction, contribute to increased inefficiency, risk and instability in the sector. The main factors are related to the high number of accidents, structural precariousness, fleet obsolescence, low-skilled drivers and high rates of greenhouse gas emissions. This paper evaluates the influence of implementing a training and feedback procedure associated with event data recorder (EDR) systems for the promotion of better behavior among professional drivers based on safety, operation and economy criteria. The analyses are based on interventions that were carried out during four monitoring phases using data generated by vehicles collected over 13 months of research. The data were converted into indicators and evaluated individually against the criteria and through data envelopment analysis (DEA). The analyses led to the conclusions that the use of EDR systems had positive impacts on all three of the criteria under analysis, and that safety levels can be increased without having to reduce productivity or increase fuel consumption. However, the safety criterion was more sensitive to the association between the technology and training process applied, leading to significant reductions in the indicators analyzed. The study contributes to the association between the methods of analysis and the adoption of specific indicators derived from time variables, leading to the conclusion that the use of EDR systems associated with management training and monitoring procedures can improve economic and operational results in road freight transport (RFT). Furthermore, using the trip data as a structural basis for the training and feedback proved to be very promising for the reduction of unsafe behavior to avoid road accidents.