Solution Extension Partnership
Production Planning and Scheduling
Performance
by Design
Avatar – powered by next-generation technologies
Designed to unite the best of two worlds: advanced planning including optimization and scheduling on record basis in combination with ML/AI-enabled decision automation and execution.
Digital Supply Chain Transformation
Unlimited
Performance
The underlying in-memory hybrid graph data model enables highest planning performance by design.
Embedded
ML & AI
Pre-built data science skills as well as complete machine learning applications for smart decision-making are ready-to-use.
Extensibility
by Design
High-performance extensibility for the data model and application functions through a unique code-to-data approach.
Unified
Core Model
One single data representation to feed different use cases: planning algorithms, machine learning skills, and optimization.
Dynamic
KPI Framework
Integrated operational reporting powered by pre-configured and freely adaptable KPIs provides holistic end-to-end supply chain.
Self
Learning
Autonomous supply chain planning and self-learning parameterization of the value chain are powered by ML/AI algorithms from Avatar’s Cognitive Engine.
One Model to
Rule Them All
The Avatar Unified Core Model (UCM) is the central data model that is used for three engines: the Planning Engine, the Cognitive Engine, and the Optimizer Engine – creating a common core for all.
The UCM enables elastic aggregation and disaggregation of data and decisions to ensure consistency: from strategic network design and tactical planning to operational scheduling – by using one and the same data structure and set.
This allows switching between different engines and solution approaches for solving today’s end-to-end supply chain challenges. Distribution network optimization can be directly combined with advanced scheduling heuristics, where unresolved overloads or order lateness can be leveled by a pre-trained ML model in one go – thereby creating superior planning results and enabling real-time decision-making.
Power of the Unified Core
Zero-Latency Planning
With its in-memory graph database, the Avatar planning engine provides high-performance computing capabilities allowing for highspeed aggregated and disaggregated scenario planning and simulation.
Data and application logic are merged into the same in-memory graph model enabling zero-latency planning – even for non-decomposable problems in complex supply chains.
In combination with event-driven data streaming, this allows for real-time scenario synchronization across users and cross-systems.
ML-Driven
Avatar’s Cognitive Engine is a key enabler for automated supply chain planning using machine learning and AI algorithms. For this it provides pre-packaged skills that are elastically linked to the Avatar Planning Engine, delivering seamless interaction with the planning algorithms while avoiding process or system rupture.
The Cognitive Engine delivers by design the technologies to automatically perform the decomposition of end-to-end supply chain problems, and it allows parallel processing of disjunct problems on various algorithms.
Optimization by Default
The Optimizer Engine provides an adaptable framework for various state-of-the-art solvers and models. End-to-end supply chain problems can thus be fixed by composing the solvers and specific models in an optimal way using MILP, Probabilistic Optimization, or even Quantum optimization in the near future.
The Optimizer Engine offers a library of models to solve standard supply chain decision problems such as strategic capability and network design or supply and distribution planning. The Unified Core Model permits flexible adjustments and extension of the standard models, allowing for a customer-specific objective function. Customers can choose between different solvers to achieve maximum performance for their individual mathematical problem.
In addition to standard deterministic formulations, the Optimizer Engine allows for covering uncertain future developments by probabilistic optimization. Related solutions offer resilient and dependable decision support for decision-makers. New customer-specific challenges can be overcome via individual and customer-tailored model formulation, executed as explorative use cases.
Power of the Unified Core
Zero-Latency Planning
With its in-memory graph database, the Avatar planning engine provides high-performance computing capabilities allowing for highspeed aggregated and disaggregated scenario planning and simulation.
Data and application logic are merged into the same in-memory graph model enabling zero-latency planning – even for non-decomposable problems in complex supply chains.
In combination with event-driven data streaming, this allows for real-time scenario synchronization across users and cross-systems.
ML-Driven
Avatar’s Cognitive Engine is a key enabler for automated supply chain planning using machine learning and AI algorithms. For this it provides pre-packaged skills that are elastically linked to the Avatar Planning Engine, delivering seamless interaction with the planning algorithms while avoiding process or system rupture.
The Cognitive Engine delivers by design the technologies to automatically perform the decomposition of end-to-end supply chain problems, and it allows parallel processing of disjunct problems on various algorithms.
Optimization by Default
The Optimizer Engine provides an adaptable framework for various state-of-the-art solvers and models. End-to-end supply chain problems can thus be fixed by composing the solvers and specific models in an optimal way using MILP, Probabilistic Optimization, or even Quantum optimization.
The Optimizer Engine offers a library of models to solve standard supply chain decision problems such as strategic capability and network design or supply and distribution planning. The Unified Core Model permits flexible adjustments and extension of the standard models, allowing for a customer-specific objective function. Customers can choose between different solvers to achieve maximum performance for their individual mathematical problem.
In addition to standard deterministic formulations, the Optimizer Engine allows for covering uncertain future developments by probabilistic optimization. Related solutions offer resilient and dependable decision support for decision-makers. New customer-specific challenges can be overcome via individual and customer-tailored model formulation, executed as explorative use cases.
Compose Your Supply
Chain Features
The Avatar Innovation Suite comes with ready-to-use, pre-built data science skills from the Avatar Cognitive Engine, as well as entire machine learning applications for smart decision-making. In addition, customer data engineers can create innovations on the platform to quickly adapt to customer specifics.
The Innovation Suite provides a freedom-of-methods approach, supporting optimizers, machine learning algorithms, simulations, heuristics, and external services.
It offers ultimate extensibility for future functional requirements in the fields of data, logistics, and supply chain.
Insight
Driven
Role-based dashboards create holistic insights into the end-to-end supply chain, supported by pre-configured KPIs that can easily be enhanced by aggregation and disaggregation, custom calculations, and enrichments.
KPI widgets can be built from scratch with Analytics Engine. Users are free to choose their preferred visualization, the level of detail, the data source, and planning actions.
The widgets can either be assembled into dashboards to create a central control tower for planning, or they can be included in any app to support comprehensive KPI-driven planning.
High-Performance Extensibility
Avatar offers unique high-performance extensibility for the data model and functions by applying an innovative code-to-data approach in the cloud.
Avatar’s integrated development environment enables customers to continuously develop and personalize their functionalities.
Avatar customers are enabled to enhance the existing standard application logic. They can easily add their own optimization models and machine learning skills or create custom planning algorithms from scratch.
Get Convinced
Get to know the world-wide first supply chain planning solution to infuse resilience, intelligence, and excellence into your operations with zero latency on all levels.