Technology
Superior by Science!
Unique Capabilities
Powerful Features
MangoGem APS Optimizer differs from traditionnal single project planning software in than in can handle multiple and complex projects at the same time, automatically sharing the available resources between the projects. In this context, projects can be anything from production orders, events to organize, construction works, ... Projects also do not need to be sequential : graphs of dependencies or time constraints are easily modeled.
Superior by Science!
If you want to improve performance, improve efficiency and reduce costs, MangoGem APS Optimizer is for you. If you want to migrate from messy spreadsheets, or are stuck with hard to implement complex programs, MangoGem APS Optimizer is also for you.
MangoGem APS Optimizer is a powerful Advanced Planning and Scheduling system including a versatile scheduling optimizer. MangoGem APS Optimizer is simple to use but can handle large and complex problems in areas such as multiple projects planning, manufacturing, event organization, routing, maintenance, staff assignment, software development, ...
MangoGem APS Optimizer is innovative, using the most recent advances in the science of operations research to tackle challenging real-life domain-specific cases and generic problems.
Unique Capabilities
MangoGem APS Optimizer is a multi-resource scheduler. Most schedulers can only handle one type of resource, or only a single resource per activity. MangoGem APS Optimizer can handle multiple instances of equipment, people, consumables, tools, locations ... and can also combine several resources together. Any activity can have multiple modes to closely model reality.
MangoGem APS Optimizer is also a multi-objective and multi-criteria scheduler. It does not just optimize one single Key Performance Indicator (KPI) at a time, e.g. minimizing makespan but with many setups or bad resource utilization, but can combine several KPIs together, produce Pareto–optimal schedules and analyze objective trade-offs. Many objectives can be selected together such as makespan, deadline satisfaction, Just In Time, resource utilization, consumables usage, activity switches, cycle time, ... MangoGem APS Optimizer can also help you analyze « what if » scenarios without effort.
Powerful Features
MangoGem APS Optimizer does not use a single “one size fits all” solver method, such as simulation, simple dispatch rules or a search heuristic.
MangoGem APS Optimizer includes many solvers and heuristics and, depending on an analysis of the case at hand, it will try and apply many methods to find the one that produces the best results. MangoGem APS Optimizer includes sequencing, heuristic search, genetic algorithms, machine learning, ...
MangoGem APS Optimizer can be used offline, online and even in real-time. It can handle large cases, with dozens of resources, and hundreds or even thousands of activities. In order to create close-to-optimal solutions very fast, MangoGem APS Optimizer can also used parallel multi- threaded solvers, and thus benefit from modern multi-core computing hardware.
The MangoGem APS Optimizer is built upon a proprietary versatile, generic and extensible manufacturing and supply chains framework model refined to be able to model accurately a large variety of use cases with the most complex requirements while being part of a standard software product. This unique design ensures compatibility with all kinds of SCM, ERP or MES systems and eases integration time and effort.
The model was designed as the cornerstone of the architecture to support different applications, features and hybridization of solvers in an efficient and highly performant way. The model handles any kind of challenge, such as attribute-based sequence setups, cleaning-in-process, tanks planning, concurrent resources, batching, time bound sequences, multiple BOM levels, order pegging, etc. The system also combines planning & scheduling into the common model to support end-to-end integrated business planning.
The MangoGem APS Optimizer uses a library of multiple solvers and multiple meta-heuristics with specifically designed algorithms that are hybridized to work on the common model. With solver heuristics notoriously difficult to fine tune, the MangoGem APS Optimizer has problem-awareness and can discover which heuristic settings work best, check validity and detect trends in data to improve ramp-up and adoption. It can also discover the most promising improvement scenarios and propose them as potential solutions to human planners. It uses simulation, deep search, meta-heuristics including Genetic Algorithms (GA), Taboo Search (TS), Simulated Annealing (SA), Swarm Intelligence (SI) amongst others. The use of AI-assisted autonomous machine learning also eases the integration and reduces the cost of implementation within overall supply chain management (SCM) strategies by making modeling easier, improving data quality and decreasing dependency on human expertise.
The MangoGem APS Optimizer incorporates machine learning Artificial Intelligence (AI) capabilities for autonomous planning and schedule optimization to be able to perform in semi-automated or even as a fully automated real-time scheduling service. But unlike the common “big data” AIs that require large historical data sets to learn, the MangoGem APS Optimizer is a small data combinatorial AI that can create its own huge problem exploration space with limited inputs.
The MangoGem APS Optimizer is platform-agnostic, running either on premises or on the cloud, on MS Windows, Mac OS and a variety of UNIX and Linux platforms. It can be used as a standalone application, or it can be easily integrated with other Enterprise Resource Planning (ERP) and Manufacturing Execution Systems (MES) tools. It can be used offline, online and even in real-time and its parallel, multi-threaded solvers ensure that it creates optimal scheduling solutions extremely fast.
The AI-assisted disruptive technology delivered in the MangoGem APS Optimizer is capable of solving many of the operational performance challenges of Industry 4.0, It uses AI-assisted autonomous machine learning techniques with multiple solvers and meta-heuristics and, depending on an analysis of the case at hand, will apply many methods to find the one that produces the best results.