Making FMCG move faster
Success in the fast-moving consumer goods (FMCG) environment is sensitive to issues such as order size and frequency, traffic congestion and vehicle use. GAVIN MYERS explores an innovative modelling technique that could turn these hurdles into measurable gains.
According to Quinton van Heerden – an industrial engineer and researcher in spatial planning and systems, at the Centre for Scientific and Industrial Research (CSIR), for commercial vehicle transport modelling – commercial vehicle movements are the result of decisions and interactions between different stakeholders in the supply chain. It is these decisions that affect the efficiency of the supply chain and the eventual costs of getting the goods to the end consumer.
Between 2011 and 2013, says the 2013 CSIR State of Logistics survey, transport accounted for 61 percent of the costs of logistics – with 88 percent of freight in South Africa being moved by road. It stands to reason, therefore, that increased efficiencies should help drive down costs.
At the recent Southern African Transport Conference, Van Heerden presented a new modelling system that he has been developing, for the FMCG sector, which aims to do just that. The system is based on a case study conducted with an actual transporter in the Nelson Mandela Bay region.
“Freight vehicles are considered a minority road user group in transport models, even though their impact on pavement damage and emissions is disproportionately large. Consequently, decisions to fund multi-billion rand projects are often based on sub-optimal models that are not really representative of the movements of freight vehicles,” says Van Heerden, noting that collaborative planning, forecasting, Just-In-Time deliveries and economies of scale all influence the frequency with which vehicles move and when orders are delivered.
“We want to understand the stakeholder interaction in a supply chain, and capture this behaviour in freight models that are more representative of freight movements,” he explains.
This form of modelling (also currently being developed in Germany) has multiple benefits; including increased efficiencies for the transporter through far more detailed analysis of its business and vehicle movements. It also allows government to make better decisions for infrastructure spend.
For the case study, Van Heerden obtained the subject transporter’s distribution data for analysis. The receivers (customers) included in the study were situated around the Eastern Cape.
In order to run the simulations, the sophisticated software required further information including a road network, agents (public transport, private and freight agents – including carriers and receivers) and configuration data with specific values, such as traffic counts. “We were looking at total tonne-kilometres moved, operating costs, travel time and distance details as well as which vehicles delivered which products to which customers,” Van Heerden says.
“We took one product and looked at the behaviour of two different customers in terms of order frequency (number of days between orders),” he explains. “One customer typically ordered twice a week, the other once a week. Order sizes also differed between the two customers, with quite big orders from the one that was further away. You’d expect this to influence the type of vehicle used by the carrier.”
Orders were dispatched from the transporter’s distribution centre using a range of different sized trucks (each with its own fixed and variable operating costs), operating during the 06:00 to 09:00 timeframe. To begin with, the carrier planned routing and scheduling for the specific day.
“Traffic congestion, for instance, has an effect on vehicle utilisation,” says Van Heerden. “We modelled three scenarios; between 07:00 and 09:00 and 16:00 and 18:00, reducing free speed to 60, 20 and 15 km/h.”
The carrier planned its routing and scheduling accordingly, using nine vehicles, including one of each type. In the case of the 15 km/h limit, the fleet couldn’t deliver all the products in time – as congestion increased, more vehicles were needed to be able to deliver all the orders. Interestingly, total distance covered did not differ much.
Customer order frequency was then considered to determine its impact on fleet utilisation. Although it stands to reason, it was proved that the more frequent orders were of a smaller size – the transporter thus used smaller and fewer vehicles to fulfil these orders. “When the order frequency was lower, however, customers ordered larger quantities and the total order size increased by almost three tonnes!” Van Heerden exclaims.
With this basic case study showing that a transporter’s behaviour is sensitive to both traffic congestion and changes by the receiver, modelling could help the transporters to plan around these factors.
“They could introduce another distribution centre to service other customers, depart earlier, take different routes or use other modes of transport,” says Van Heerden. “The process should be repeated until all iterations are complete and the results can then be analysed.”
Finding the right balance will allow transporters to reduce the time and cost of delivering goods and modelling might even allow transport and road planners to make better decisions in key areas – at the end of the day saving the consumer from footing the bill.