Cargo containers have emerged as a fundamental means of transporting manufactured goods in a supply chain logistical system. The need for containers arose when the movement of individual boxes, barrels and crates became time-consuming and labour intensive at ports (Northport 2009). The bulk transportation of goods resulted in the introduction of containers in the 1950s, followed shortly thereafter by the building of containerships (ExportHelp 2010). This promoted intermodalism in which a container is moved via various modes of transportation in a simpler, logistical process, hence expanding trade amongst countries (Northport 2009). This logistical process has improved tremendously since the early era, when cargo ships spent more time loading and dispatching shipments at ports than on actual transportation via sea (Cudahy 2006). The Durban Container Terminal has the largest capacity to handle containers in the southern Hemisphere (KZN Department of Transport 2011). It was established in 1977 and moved approximately 78 493 containers in its first year of operations (Transnet 2011a). Currently, the terminal has six berths and handles 1 474 978 containers a year, including various types of containers, from International Organisations for Standards (ISO) containers of general purpose to abnormal containers (Transnet 2011a). These container standards were institutionalised in the 1960s with the growth of intermodalism in mind in order to achieve commonality amongst freight and facilitate movement across different modes (Coyle et al. 2011). Containerisation contributes to a higher efficiency in the development of intermodal and multimodal transport operations.

The movement of freight in containers increases the efficient handling and storage of products using containers ranging from ISO to technologically improved smart containers. The network theory views any system as a set of interrelated actors or nodes (Tate, Ellram & Golgeci 2013:266), whereby the actors can represent entities at various levels of collectivity and linkage, such as persons, firms, countries and other participants in the network (Borgatti & Li 2009:2). Intermodal networks represent an important way to diffuse ideas and practices amongst road freight transport operators, freight forwarders, intermodal supply chain customers and the Port of Durban within the network. Network theory enhances the understanding of the role of supply chain logistics partners, governmental organisations and Port of Durban on both inbound and outbound intermodal transport networks. Although the driving forces behind intermodal economic practices include government regulation, customer requests and costs savings (Rugman & Verbeke 1998:363–375), collaboration in a supply chain network is required for extended enterprises to share knowledge and skills on consolidated diffusion of intermodal economic practices and Port of Durban operational process initiatives (Pagell, Wu & Murthy 2007:133–143).

Diffusion of innovation is influenced by leadership attitude toward change, interconnectedness and system openness within the context of technology (Rogers 2003) and top management support within the organisation context (Oliveira, Thomas & Espadanal 2014:497). The technology–organisation–environment framework (Tornatzky & Fleischer 1990) assists in providing a more comprehensive perspective for understanding Transnet's central IT adoption (Navis SPARCS N4) (Transnet 2011b) by including the technology, organisation and environment contexts for the inbound and outbound Port of Durban intermodal transport network (Zhu et al. 2006:601–616). If the diffusion is the process in which an innovation is communicated over time through certain channels amongst members of a social system (Rogers 2003), the principal characteristics of the Transnet's central supply chain IT system are applied to the planning of road freight transport infrastructure and positive impacts for intermodal transport management, particularly planned Port operations processes and controlled traffic congestion in surrounding Port of Durban.

Arguably, the adoption theory, points out that many innovations do not achieve the expected results for failure to satisfy the requirements of potential adopters (Figueiredo 2005) relating to an efficient intermodal freight logistics chain. The dynamic capability theory elaborates more on the principles of how the Port of Durban and intermodal freight logistics partners can integrate, build or deploy and reconfigure their internal resources and external competencies in changing the containerisation environment (Newbert 2007; Teece, Pisano & Shuen 1997). The dynamic capabilities are critical for performance in the intermodal environments as they determine the flexibility, compatibility, clockspeed and delivery efficiency (Teece 2007). According to the institutional theory, external isomorphic pressures from competitors (Ports of Richards Bay, Port Elizabeth and Cape Town), trading partners (intermodal and multimodal transport services providers, freight forwarders and bonded warehouse) and customers (all Transnet Port Terminal customers) likely induce forms to engage in certain supply chain logistics management practices (Sodero, Ravinovich & Sinha 2013). This unique link in the logistics process networks provides direct availability to clients and allows for long haulage.

Road transport is described as the most effective and flexible mode of transportation in South Africa; it has overtaken rail transportation and dominates the freight movement because of the convenience and services it offers. The road transport network offers faster delivery, cheaper packing systems and shorter lead times. As a result of containerisation and the road transportation network, South Africa faces the challenge of the carriage of containers that exceed the gross mass weight laid down by legislation (Department of Transport 2011). This leads to increased liability costs and infrastructural damage to major roads and national highways, as well as being the primary cause of many road accidents. Poor road conditions can result in higher costs and litigation, impacting containerisation, the intermodal activities and the road transportation network (Miller 2001; Teravaninthorn & Raballand 2009).

The increased volume of road freight networks has aggravated traffic congestion surrounding the Port and storage depot areas. The traffic congestion is the result of the uneven split between road and rail transportation (Council for Scientific and Industrial Research 2010). Whilst 88.7% of containerised freight is transported via road, only 11.3% travels via rail (Council for Scientific and Industrial Research 2010). To address road congestion, the KwaZulu-Natal Department of Transport implemented a system whereby trucks enter the Port terminals based on scheduled appointments with the freight forwarder (KZN Department of Transport 2011). Freight forwarders act as fourth-party service providers in transporting containers from shipment through the road transport network and they are fully liable for the freight in their possession from its origins in ports to the destination of final customers (Cassidy 2010). Contracts between suppliers of containerised freight and the final user need to be prioritised in order to avoid liability claims and legal issues (Goga 2012).

Efficient container management is required to maintain safe and smooth operations; radio-frequency identification is able to trace a container's location at a terminal, ensure the maintenance of vehicles delivering containers through the road network, calculate the loading of the container and ascertain any problems encountered with containers at terminals (Barro-Torres et al. 2010). It enhances the security of containers (container identification, determination of seal breakages on container) (ISO 2009) and improves efficiency at ports (Barro-Torres et al. 2010). Dedicated short-range communication allows for vehicle traffic to be monitored at portside (Coronado Mondragon et al. 2012). This wireless network facilitates the movement and coordination of road transport networks in and around ports through communication using improved logistical applications. Scanners and X-ray technology are used during inspection and customs clearance at ports and terminals to determine hazardous substances, whilst an electronic seal is attached to the door of the container (Roberti 2005) and is used to detect when containers’ doors have been opened by unauthorised personnel. Security aspects of the supply chain consist of a series of layers that promote the safe delivery of containers which have not been subject to theft and mishandling (Raymond 2006).

Although wireless technology promotes the advancement of tags that assist with the tracking and tracing of containers (ISO 2009), Transnet's central system provides Transnet Port Terminals (TPT) customers with a seamless and integrated service across the ports. The TPT became the port operator to the Navis SPARCS N4 terminal operating system from a central location across 21 marine and rail terminals. Transnet's central system affords TPT customers with decreasing costs of doing business and increasing competitiveness within South Africa through a more effectively integrated supply chain (Ngcobo 2009:1). If the web-based SPARCS N4 operating system governs the movement of all container logistics and operations, the improved intermodal logistical system support should lower the supply chain operating costs through greater control over resources and intelligently respond to changes or disruptions in the flow of cargo. Havenga, Simpson and De Bod (2012:38) highlight that national freight logistics costs are high, with road infrastructure challenged by increasing concern of environmental impact, and the authors believe that intermodal solutions could reduce transport costs. Scheinberg (2009:2) adds that Navis allows ‘inter-terminal integration and synchronised planning linked in the chain to track scheduled vessel progress and plan efficiently across several terminals in the country’. Musa, Gunasekaran and Yusuf (2014:176) assist in describing tracking as the determination of the identity and state of a product, whilst tracing is described as inference to the product's path and history from downstream to upstream of the supply chain.

This end-to-end supply chain visibility by tracking and tracing reflects a means of security and process control and optimisation in multisite management systems. The central system deems to offer a certain level of flexibility and depth of effective functionality to abate the challenges of containerisation on intermodalism in a sea–road freight network. The centralised visibility of real-time information between TPT operations and cluster of logistics customers improves supply chain intermodal logistical systems and allows streamlining operations to realise the benefits of multisite management (Port Technology 2014). In a competitive environment, the profitability of a focal company such as Transnet depends strongly on its ability to manage complex relationships with its business partners, especially the freight forwarders (Caridi et al. 2010:593). The focal company is identified by stakeholders, including customers, as being responsible for the specific modes, products or services and the coordination and integration of physical and technological resources and centrally synchronised real-time information flows (Caridi et al. 2014:1).

The first hypothesis set was as follows:

Hypothesis 1 _A : Decision-making in the various fields of employment in this sector has made a positive contribution to efficiently synchronised-quasi-real-time delivery schedules.

The contingency table generated a relationship between bivariate-based variables to determine the appropriate associations from 110 valid cases at significance level (0.05).

Table 2 depicts the relationship between the strategic optimisation factors (delivery time, peak time and night transport – Table 2a) and fields of employment (intermodal, fourth-party, road freight, freight forwarder and depot – Table 2b) on the underlying magnitude of agreement amongst the statements. The overwhelming majority (80.9%) of the respondents concurred that the quasi-real-time delivery schedules should be properly and efficiently synchronised across enterprises with freight forwarders. Furthermore, 71.2% agreed that stringent rules should be applied to road freight transport operations during peak times in order to alleviate traffic congestion on the road networks. In terms of the percentage agreement amongst the fields of employment, depot (35.5%) and road freight (31.8%) players represent a combined majority of the respondents (67.3%) that tentatively supported closer synchronisation amongst delivery partners, night operations and restricted operations during peak times. Optimal delivery times, restricted operations during peak times and night transport operations were moderately supported by intermodal (9.0%), fourth-party (17.3%) and freight forwarders (6.4%). The tests generated significance levels less than 0.05 (degree of freedom = 1; 16), which failed to accept the null hypotheses. This study therefore infers that the proper alignment of quasi-real-time delivery schedules requires a closer synchronisation with freight forwarders in order to enhance logistical supply chain performance in road freight networks. In order to alleviate road freight congestion, the study points to night transport operations and operational restrictions during peak times.

The second hypothesis set was as follows:

Hypothesis 2 _A : The road freight transport mode has an essential relationship with either delivery time schedules, or transport routing, peak times, full-truck load lane, a reduction in the number of trucks or night transportation.

Cohen et al. (2003) describe multiple regression as a flexible method of data analysis that may be appropriate whenever a quantitative criterion variable is to be examined in relation to any other factors expressed as predictor variables. The Pearson correlation matrix indicated an acceptable correlation between all possible pairs of variables and a significant positive correlation (p < 0.05) with a dependent measure (transportation mode). The criterion variable was positively correlated to delivery schedules, routing traffic, peak time, full-truck load, reduction in trucks, night transport networks inspections and risk reduction, advanced technologies, security protocols, prolong delivery times and proper documentation when using the 5-point Likert scale. Only two predictor variables (prolong delivery and full-truck load lane) were entered into prediction model 2 after stepwise procedure with a multiple R of 0.289 and both prolong delivery and full-truck load lane were significantly entered into the regression equation.

The multiple regression technique was chosen to hypothetically determine the relationship between the dependent variable (transport mode – road freight transport network) and predictor variables. Table 3 explains only 8.3% of the variance (improved from 4.2% in model 1) in the road freight transport mode. The second model produced two dimensions; prolonging delivery ( β = -0.25, p = 0.008) and a full-truck load lane ( β = 0.209, p = 0.031) at significance level of 0.05 were found to be considerably related with the road freight transport mode. In terms of an efficient road freight transport mode to alleviate road congestion, this is predictably related to full-truck loads being allocated a separate lane and further associated with prolonging delivery times to customers.

These results link the road freight transport mode and the challenges of containerisation as determined by regulated road freight infrastructural design for full-truck loads and the efficient customs clearance procedures to enhance the on-time delivery service level to customers. Two models from the stepwise procedure were produced by regression analysis without concern for the multicollinearity problem, with both the variance inflation factor below 10 and tolerance more than 0.10, reflecting acceptable positions. In the case of normality, the scatter plot maintained points between the ranges of -3.3 and +3.3 without any outliers. The output stated the measure (min. = 0.001, max. = 0.149) of leverage on how much an observation influences the regression coefficient and this study revealed acceptable hat values within the rule of thumb (leverage goes from 0 to 1). In examining the degree of autocorrelation, model 2 indicated the value (2.036) of the Durbin-Watson statistic to be consistent with the ideal range values (between 1.5 and 2.5), with no problems related to autocorrelation.

The third and final hypothesis set was as follows:

Hypothesis 3 _A : Supply chain technological advancements in container-based road freight transport networks produce mean differences towards either containership theft, or in-transit communication, systematic risk or inbound and outbound coordinated scheduling.

The ANOVA uses a single-factor, fixed-effects model to compare the effects of one treatment on a continuous dependent variable. Table 4 shows the ANOVA tests that were conducted between advanced technologies and both containership theft and in-transit communication. The Levene's test of homogeneity of variance was greater than the level of significance (0.05); hence, there was no reason for further robust tests of equality of means as the assumption had not been violated. The underlying variables between groups showed significance levels of all greater than 0.05 and the Turkey test depicted the means without difference to each other; thus, there was no significant difference observed through the means tests between advanced technologies and both containership theft and in-transit communication. The second component of the set investigated whether the improvement in technology mitigates the systematic risk in transportation whilst enhancing the scheduling of container movement on the basis of advances in technology. Both systematic risk and inbound and outbound scheduling indicated significant difference amongst the scores of means, with F = 3.649, p = 0.15 and F = 4.255, p = 0.007, respectively. Technological advancement can tentatively provide significant mean differences in mitigating systematic risk in the road freight transportation networks and closely synchronised inbound and outbound scheduling systems.