Skip to content
Opt Dir

Glossary ยท approach

ATM Cash Management

OR problem of forecasting each ATM's cash demand, deciding replenishment timing and amount, and balancing the immobilisation cost (interest opportunity + security risk) of idle in-ATM cash against the stock-out cost (customer complaints + lost transactions) of an empty ATM; joins forecasting + stochastic inventory + cash-in-transit (CIT) routing.

ATM Cash ReplenishmentBank Cash ManagementATM Network Cash Optimisation
ATM Cash Management is the operations research problem of a multi-ATM commercial or participation bank: for an ATM network (typically 100-3,000 machines) jointly decide (a) each ATM's daily opening cash quantity (by denomination โ€” mix of small/medium/large notes), (b) which ATM is replenished how often (daily, every 2-3 days, weekly), and (c) the nightly route of the armoured cash-in-transit (CIT) vehicle. The balance is between **immobilisation cost** (5-15% annual reference interest opportunity on idle in-ATM cash + insurance premium + security risk) and **stock-out cost** (empty-ATM complaints โ€” lost transaction fee + brand-reputation damage). It is distinct from classical IRP (#048) because denomination mix, security/insurance cost, location-specific demand patterns (shopping centre weekend, campus month-end, office payday) and banking-regulator CIT-monitoring regulation are cash-specific dimensions. Foundational work is Simutis-Dilijonas-Bastina-Friman-Drobinov (2008) on ATM-network optimisation; the canonical stochastic formulation is Castro (2009); the modern survey is Wagner-Bachner (2018); group-demand forecasting is Ekinci-Lu-Duman (2015); the integrated inventory + distribution VMND algorithm is Larrain-Coelho-Cataldo (2017). The service level ฮฒ = 0.95-0.99 is priced as a chance constraint; rolling-horizon re-planning can refresh the solution intra-night with telemetry data.
ร–rnek

A mid-size commercial bank with 1,200 ATMs runs an integrated demand forecasting (Holt-Winters seasonal + month-end/payday/holiday anomaly flags) + stochastic inventory (ฮฒ = 0.97) + CIT VRP (22:00-06:00 night shift) solver on a 14-week pilot in a single city covering 180 ATMs, in parallel with the existing manual solution; immobilisation cost drops 22%, empty-ATM events fall 47%, CIT vehicle productivity rises 13%.

Where this term appears

Esc Close