(477b) Modeling Joint Performance of Financial Budgets and Operative Plans in Supply Chains

The hierarchical production planning (HPP), the classical procedure applied in supply chains (SC), decouples the planning and scheduling decisions from budgeting considerations, and decides first on planning and scheduling and then it simply performs the corresponding financial computations for budgeting. As opposed to this approach, here we show that one can obtain significant financial gains by considering both, plan and budget problems, simultaneously in a financial and operative simulation model. We present a simulation descriptive model in which the supply chain operation has a liquidity control (cash is retained at a safety level), while are tried the best options to invest the surplus if opportunities exist. To formalize the liquidity status in the financial model, are used the typical financial categories of the balance sheet, that is, the current assets (CA), fixed assets (FA), current liabilities (CL), long term debt (LD) and equity (E). The mathematical balance equation (CA+FA-CL?LD-E=0) controls the net flows of the firm's cash, making it possible to control the updated liquidity in the integrated functional model. In this joint environment the movements of events related to financial supply chain activities that create the monetary flows, are freely permitted. Therefore, It is feasible to flatten the money outflow peaks moving production tasks in order to achieve the optimal earnings with synchronization of cash flows. On the budget side we manage simultaneously decisions concerning cash management, compensating balances, prompt/delayed payment of sales/liabilities, investments in marketable securities, credit line, loans, repayment policy, pledging, manpower wages, debts, retrofitting, new assets for bottleneck solution, etc. (Badell et al., 2005; Romero et al. 2003, and Guillén et al., 2005). In this work we investigate several solution techniques with different performance measures in order to create a common modeling building block capable to represent operative and financial operations. Our objective is the development of the necessary scheduling concepts to remake the comprehensive understanding that should ultimately facilitate the modeling of complicated systems considering financial operations in the same manner as the operative tasks. In order to understand completely the behaviour of a complex system, it is vital to understand the workings of its components to detect its analogies and differences and the key management regulation loops regarding to the enterprise functionality. The theory of scheduling uses different types of methodologies like combinatorial procedures, simulation, graph techniques, network methods and heuristic approaches for addressing efficiently sequencing problems. However, the assignment problem has always been calculated in one dimension (t.e. timing units) when searching the best position of batch sequences. Historically, engineering computer aided scheduling tools explicitly represented in Gantt charts the resource time, gathering directly or indirectly the whole measurement responsibility, and forgetting its omnipotent money partner in off, left slave of previous production decisions, disoriented about the operative routine and practice and hence, mistaken during planning. Due to the fact that chief financial managers (CFO) basically need information about the batches interferences in plant during planning, the batch assignment in a scheduling simulation tool for financial planners can be represented as a black box during computations, ignoring the internal technological details, and recovering these details subsequently, if required (information related to unit operations is available). Regarding methodologies, and with the intention of making easier the modelling task, we developed a normative assignment/sequencing solution using Symmetric/Asymmetric Traveling Salesman Problem (TSP/ATSP) formulation that uses the overlapping times (or not_overlapping times) between batches with economic weight in order to fulfill the scheduling task [Badell, Fernández and Puigjaner, 2000]. Schedules are generally evaluated by aggregate quantities that involve information about all jobs, resulting in one dimensional performance measures. Measures of schedule performance are usually function of the set of completion times in a schedule. The aggregate performance measures defined induced others, as makespan, mean flowtime, mean tardiness, maximum flowtime, maximum tardiness or the number of tardy jobs, being all of them a function of the set of job/plan completion times. These performance measures were denominated regular due to the fact that when the scheduling objective is to minimize Z, Z can increase only if at least one of the completion times in the schedule increases. These indicators played the role of intermediaries linking time-to-technology-to-cost assuming that, for example, minimum makespan can be extrapolated as maximum profit, being this not always true. Moreover, research is almost entirely dedicated to the study of regular intermediate performance measures. Besides mathematical exact solutions, limited to a small number of products owing to its high solution time, we also adapted financial and operative scheduling tasks to operate jointly in professional advanced planning and scheduling systems (APS). The APS approach relies on the creation of computer models that permit to schedule resources assuming finite capacity. Besides the exact solutions, APS systems can also work with heuristic procedures. In chemical industry the recipes give the time guidelines for production. To represent the financial tasks we used financial ?recipes? and created fictitious units that compile the financial issues in the same manner as production tasks do. Therefore, Gantt charts can depict financial operations joining together with the production operations representation. Moreover, there is no reason for not applying to finantial shedules the same achievements of production scheduling in Gantt charts: they can be handled interactively, they can be displayed with a desired granularity shown in clicked windows, and can even be interactively rescheduled. So plans and budgets can be easily modified, as well as the resulting profiles of financial resources like money or balance sheets. At present, it is not enough to integrate the SC measuring information of its performance with financial key performance indicators (KPI) as profit, costs, earnings, value added, etc. In this position KPIs assume a passive informative role. These optimal financial figures obtained must go ahead determining a list of optimal enterprise decisions. Where to place the money on hand, also requires optimization techniques and today these decisions are determined freely, based in intuition, while liquidity preservation is a great unknown variable for all the staff. Here we measure the performance by means of a multidimensional aggregate quantity that jointly takes into account the timing of process, the cleaning tasks for the respective batch sequences, and the economic/financial value added during production in the SC. This permits to decide simultaneously the dominance of the most valuable assignments as a function of time & money. This time-money aggregate can be obtained calculating the ?density? of profit per hour of completion time, which obviously is schedule dependent. This density can be easily calculated as the hourly contribution to profit of the batches considered, divided by its real processing time (e.g., considering overlaps). Thus the current scheduling criteria sets priority on time when the intimate relation between these two objectives requires a time-money performance measure. We propose the creation of a simulation modeling tool to aid planners make their job, modifying the trial and error classical procedure of HPP and creating an adequate computer aided system (70% of the financial managers currently use Excel platform to make budgets). We present the results of two cases of study, the first one mainly within an enterprise scope and the second one, within a detailed supply chain scope. The methodology compares the results of the decoupled models with the integrated one.

Badell, M., J. Romero, L. Puigjaner, ?Optimal budget and cash flows during retrofitting periods in chemical process industries?, Intern. Journal of Production Economics, 95 (2005) 359-372.

Romero, J., M. Badell, M. Bagajewicz, L. Puigjaner, ?Integrating Budgeting Models into Scheduling & Planning Models for Chem.Batch Industry?, Ind. Eng. Chem. Research (2003).

Guillén, G., Badell, M., Espuña, A, Puigjaner, ?Integrating short-term budgeting into multi-site scheduling? Comp. & Chem. Eng., in process, 2004.