Seven mathematical solvers for supply chain optimization

1. Gurobi: LP, MIP, QP, NLP

✅ One of the fastest solvers for large scale LP and MIP problems
✅ Offers various parameters to fine-tune the solver for specific problems

🚫 Expensive for commercial use
🚫 While it can handle some nonlinear problems, it’s primarily optimized for linear and mixed-integer problems.

2. Knitro: NLP, MINLP

✅ Excellent for nonlinear and non-convex problems
✅ Well-suited for industries with complex production and logistics processes

🚫 Expensive
🚫 Not as efficient for discrete or combinatorial problems

3. SCIP: MIP, CP, NLP

✅ Open-source, accessible for testing and prototyping
✅ Performs well in scheduling, vehicle routing

🚫 Not as fast as Gurobi or CPLEX for large-scale MIP problems
🚫 Requires more expertise in constraint programming for complex models

4. CBC Solver: LP, MIP

✅ Free to use and open-source, making it accessible for companies with limited budgets
✅ Good community support and frequent updates

🚫 Slower than commercial solvers especially for large-scale or highly complex models
🚫 Lacks some advanced features that are available in paid solvers (e.g., advanced cut generation, parallel computing)

5. MOSEK: LP, MIP, QP, Convex NLP

✅ Excellent performance for large-scale continuous optimization.
✅ Strong support for conic and convex optimization problems.

🚫 High licensing fees can be a barrier for some users
🚫 Not suitable for non-convex problems, which can restrict its applicability in certain scenarios

6. Microsoft Excel Solver: LP, NLP, MIP

✅ Easy to use, requires no programming skills, and is widely available in many businesses.
✅ Works well for small-scale, basic supply chain problems like simple transportation and inventory optimization.

🚫 Not suitable for large or complex problems
🚫 Lacks the advanced features and customization options available in dedicated solvers.

7. Google OR-Tools: LP, MIP, Vehicle Routing, Scheduling

✅ Free to use, backed by Google, and frequently updated with new features.
✅ Excellent for vehicle routing, transportation optimization, and similar logistics problems in the supply chain.

🚫 Mostly suited for vehicle routing and scheduling; not as versatile for other supply chain problems
🚫 While it’s actively supported, the documentation is not as thorough or polished as for commercial solvers.

P.S. All these solvers need of course a modelling language & modelling environment to capture in a mathematical form the business requirements. Also, a deployment platform like Nextmv could be added.