The fourth-biggest cost in liner shipping
Bunker, terminal, charter hire — and then empty repositioning. That is the cost stack in the liner industry. Laden movements are covered by the consumer; empties are not. They are a pure operational expense, carried entirely by the carrier.
Liner operations teams are asked to meet dynamic, fluctuating cargo demand with a limited and dispersed supply of containers. Bloated stock levels, high storage costs dictated by commercial demands, high maintenance and repair costs, and a lack of clarity around future demand and supply all aggravate decisions around purchase, sale, lease and sub-lease. Most carriers manage all of this on outdated systems with manual planning.
“If liner shipping companies hope to remain competitive, operationally agile and financially stable, they need to turn to technology to guide them successfully into the future.”
What an empty-repositioning optimiser is actually meant to do
Before getting into how OptiBox is built, it is worth stating what an optimiser of this kind is supposed to deliver. Most of these are missing from a regional, spreadsheet-based process.
- A global view of imbalances and repositioning plans that prioritise the best-contributing cargo
- Reduced storage costs through a planned evacuation plan
- Direct control over repositioning costs
- Improved asset utilisation through reduced container idle time
- Identified opportunities for one-way leases with other box lessors — cutting owned inventory
- Stock inventory levels brought down from bloated to ideal
- Lower maintenance and repair (M&R) costs by routing to depots with the lowest M&R costs
- On/off-hire base costs minimised by picking the right drop-off and pick-up locations
- Faster container turnaround, increasing container velocity
- Global rather than regional repositioning decisions — which generally produces a lower total bill
How OptiBox works — four layers
OptiBox is powered by optimisation algorithms, artificial intelligence and time-series forecasting. The optimiser runs four layers, in sequence, to turn raw operational data into an executable repositioning plan.
“OptiBox is backed by SEDGE technology — data cleansing, data munging, statistical analysis, data exploration, time series forecast, seasonal trend and predictive analytics.”
- External data — n-weeks of historical records, equipment management system (EMS) routing, index costs, on-hire / off-hire data, container M&R records, depot storage records and the vessel-schedule network
- Business rules and master data — port-pair quantity, transit days, trans-shipment ports, port rules, equipment hiring limits, vessel TEU and weight constraints, transit time profiles, depot storage and handling, container repair costs
- Demand forecasting and modelling — laden and empty forecasts and a global equipment imbalance status, with safety stock defined per port and equipment type
- Optimised results — an optimiser engine that repositions surplus containers to deficit locations while satisfying every constraint, calculating estimated cost in advance and surfacing the imbalance geographically for global planning
What-if simulation, Plan vs Actuals
A plan is only as good as the conditions it was built under. OptiBox includes a what-if simulation function — users can input specific conditions, scenarios and business rules and re-run the optimiser to see the financial and equipment-availability impact of each scenario. The differences against the baseline global plan are highlighted.
Separately, a Plan vs Actuals view lets carriers compare the planned repositioning against what actually happened in the field — exposing where execution diverged from intent, and where the assumptions need updating.
Dynamic routing and cost calculation
For every surplus-to-deficit pair, OptiBox identifies all possible cross-joins, then generates the best route within the existing constraints. The total cost of moving a box from A to B is summed across every leg: on-hire and off-hire, load cost, local trucking at origin and destination, feeder cost, haulage cost, depot handling at origin and destination, discharge cost, and trans-shipment load costs.
Route generation itself draws from multiple sources: the routing master, existing haulage contracts, existing feeder contracts, hub-route creation (nearest seaport), auto-route creation (nearest inland location), and a distance route finder. Routes are appended to the base routing wherever the customer demand and proforma generation warrant it.
Built to be used, not just to be impressive
The platform underneath OptiBox is technically complex. The point of OptiBox is that none of that complexity surfaces to the user. Operations teams do not need a team of data scientists to operate it. With a few clicks, the existing team gets the benefit of the data science underneath, and the value of cloud-based AI, without skipping a beat.