IMPL is a proprietary Fortran-based modeling and solving platform for large discrete, nonlinear and dynamic industrial optimization and estimation problems. It supports both decision-automation tasks such as planning, scheduling and control optimization and data-analytics tasks such as data reconciliation, regression and parameter estimation. It is flowsheet, fundamentals and formula based, built on the Unit-Operation-Port-State Superstructure and Quantity-Logic-Quality Phenomena framework, which gives it the structure of a generalized network-flow system with setups and property pooling. This allows IMPL to represent superstructures, routings and flowsheets together with detailed process behavior across industries seeking economic, efficiency and emissions optimization.

The platform follows a small-high-payback implementation philosophy and is interfaced through ASCII files or APIs callable from languages including Fortran, C, C++, C#, Excel/VBA, Java, Python, Julia, R and MATLAB. IMPL provides unique engines such as SLPQPE, SECQPE, SORVE and SDSDRE. It also connects to major LP, QP, MILP and NLP solvers such as COINMP, GLPK, LPSOLVE, HIGHS, SCIP, CPLEX, COPT, GUROBI, MOSEK, XPRESS, LINDO, CONOPT, IPOPT and KNITRO. IMPL can also serve as a platform for hybrid or nonlinear model predictive control in real-time environments.