The AMPL modeling system has been used in thousands of business operations worldwide. Applications are found in manufacturing, transportation, energy, finance, and many other sectors where complex decisions must be made to minimize cost and maximize efficiency.
Through the use of an industry-accepted simulation approach, an advanced analysis methodology combined generation, transmission, loads, fuels, and market economics into one integrated framework to deliver location dependent market indicators, transmission system reliability and market performance indices.
AMPL x ABB built an optimization model using a mixed-integer linear solver, involving millions of variables, and tens of thousands of integer variables, with a solve time of 10 minutes.
With the C++ API, this model can be easily embedded into ABB’s GridView product. This solution was deployed to 30+ companies and hundreds of customer side-users.
One of the biggest challenges in managing a large power grid involves two decisions: When to turn power plants on and off; and how to transmit power over the grid to meet demand.
Assign each representative a similar number and quality of accounts, while giving priority to assigning higher quality accounts.
AMPL was chosen based on its ease of use, speed, reliability, and ability to handle large problems. Integrated data indicated a quality score for each account predicting revenue increase if contact by a representative, and the location of each representative. AMPL supported the maximization of total score of all assigned accounts with no more than a 15% variance between representatives in number and quality of assigned accounts.
With AMPL is embedded in Dropbox’s systems, customer data is extracted from Salesforce and customer scores are computed using the scikit-learn Python toolbox. An AMPL script reads the file of score data and results from optimization are written to an Excel spreadsheet.
At Dropbox, over 500 million users upload over 1.2 billion files per day. With tens of thousands of large business customer accounts, and hundreds of sales representatives worldwide, Dropbox has enough to cover most, but not all, of their customer accounts.
The goal was to determine all shipments to all stores to provide the greatest benefit to the company overall. We needed to determine the number of each size be shipped to each store, with the expected sales of the sizes.
Collaboration of UCLA & MIT researchers with Zara corporate technology staff led to the use of a mixed-integer linear solver with AMPL’s model-based optimization to create a forecast model approach. This replaced the legacy approach of store mangers and warehouse allocation teams…
Two years of implementing, testing and deployment led to a client app where 60 employees of the warehouse allocation team can use demand forecasts and warehouse and store inventory to create shipments.
Zara warehouses ship clothing to 1500 stores every week. The shipments of different items changes week to week for up to 3000 different clothing items and up to 8 different sizes. Decisions need to be made within hours of receiving sales data.
The goal was to generate good packing plans for a day’s orders, without using more racks than needed, and to finish in time to get orders out.
AMPL was chosen for its dramatically better solutions and selection of solvers. The optimization solution implemented the use of AMPL model, data, and scripts with minimum problem size of low 100s of thousands of variables and constraints to maximum a problem size of 100 million variables and constraints with a solve time around 10-45 minutes.
A VBA-modified spreadsheet for data prep and results reporting was developed, and one replenishment specialist uses the took multiple times a day, allowing for far fewer people, faster loading of racks and fewer trucks required.
Young’s Plant Farm grows plants of many kinds and sizes, supplying large retailers such as Lowe’s and Walmart, by packing plants on special rolling racks that are then loaded on trucks for transport.
See how we help companies achieve impactful results across multiple industries and domains
Some of the world’s largest hedge funds minimize deviation on their benchmark-tracking portfolios.
Traditional, Renewable and Storage asset operators optimize dispatch and charging at utility scale.
North American, European and Australasian grid operators optimize existing flows and plan grid expansions.
Storage Asset Owners maximize profitability in Frequency Control and Ancillary Response market participation.
Newly reshoring conglomerates remodel and optimize their supply chains fast in response to deglobalization.
Manufacturers like General Electric, Kimberly-Clark and Southern Company minimize time and cost of product assembly.
Roger’s EV optimize electric drone fusilage design for optimal aerodynamics and battery/payload balance.
Telescope and satellite operators optimize asset trajectories and orientations.
Research teams and drone logistics innovators model optimal drone routes for new delivery and service fleets.
Telcos optimize communication packet routing and signal processing across consumer telecomms networks.
Large Data Center Operators pair Machine Learning and Optimization techniques to reduce power consumption.
The US Census Bureau optimizes report creation and data anonymization routes on national scale datasets.
Boutique investment consultancies model and optimize clients’ tax exposures on existing and planned portfolios.
Economists implement very large stochastic optimization models for solving dynamic continuous-time problems.
A maker of beef patties employs an optimization model as the backbone of its planning, blending and procurement system.
Oil and Gas tanker fleet operators maximize profit and minimize time at-port.
Volkswagen optimizes placement of North American vehicle distribution centers.
A pharmaceutical and chemicals giant improves scheduling of a special toxic waste furnace.
Delta, BC Rail and other transportation majors optimize crew works schedules, hiring, training and vacation rostering.
Truck fleet operators optimize real-time pricing offered on their market platforms.
We obsess over the optimization process so you don’t have to
Our developers, owners and operators are expert modelers in the field of mathematical optimization. We listen to our customers feedback and create new solutions and offerings so you can spend more time focusing on on the solutions and less time on the problem.
Start a free, full-featured trial of AMPL for personal, academic, and commercial prototyping purposes. This program includes open-source solvers, as well as 30-day trials of commercial solvers.
Start faster with less money down. Begin a subscription on AMPL and solver licenses today.
Need a starting point? View our model repository for a free entry point into optimization.
For small and open source solver projects, use our community edition license.
At AMPL, we believe in putting people and their initiatives first. We work closely with our customers from model development to deployment, offering phone and email support, to create models and solutions that make a difference in your business day to day.
See all our license pricing options for AMPL and solvers we sell.
Download a free, full-sized, full-featured AMPL license and trial all our commercial solvers free for 30-days each.