I am your specialist for Near-Sourcing your Business Tasks regarding…

(i) the Development of AI models (Unsupervised, Supervised & Reinforcement Learning) for your Business case,
(ii) the Optimisation of specific Business Tasks (using exact models, Mixed Integer Linear Models, or Machine Learning, GRASP and other heuristics), e.g. in context of Production and Supply Chain Management processes (Ordering, (Re-)Scheduling, Planning, Order Release), and
(iii) the Implementation and Automation of your Business Workflow Processes.

I am Manuel, a self-employed worker with a PhD in Information Systems and a Master degree in computer science. I work on projects in several areas. Due to studying computer science (MSc) and information systems (MSc + PhD) I have a wide range of knowledge covering computer science and economics, like logic [2,3,6], programming and algorithm development, especially in the field of artificial intelligence [4,5,8,9] , production planning and optimization [1,7,8,10], with a focus on operations research and operations management. This is completed with the fundamental know-how on how to successfully apply the latest scientific achievements of these research areas in modern companies [11].

Throughout my PhD, see [9], I concentrated on Artificial Intelligence, with a special focus on Reinforcement Learning, in the context of production planning. Although we found that machine learning, in particular artificial neural networks are tools that can be applied in manufacturing firms [1,7], we also observed that current reinforcement learning algorithms are not suited for optimization problems arising in modern companies, especially in the context of production planning [8,9]. More precise, the widely-known algorithms are built for game-like environments, where only a sparse reward function is used, e.g. 0 reward until the end of the game. Hence, by analysing the fundamentals I developed an alternative reinforcement learning approach, overcoming the problems of the standard techniques [5,9]. This led to an algorithm that outperforms well-known standard reinforcement learning [5], where we apply it in the context of production planning [4].

Previous projects range from statistical analysis using distributed non-linear lag models (dnlm) in order to analyse lagged effects of hospitalization for sand storms in the canary islands, the health effects of the recent volcanic eruptions in the island of La Palma, to applying Reinforcement Learning in companies like Engel and Bosch, or optimization of short-term production planning tasks like capacity management and routing and task selection of automated guided vehicles in manufacturing environments. The former two projects are being conducted in conjunction with the University of Valencia (Spain) and the local government of the canary islands, while the machine learning projects are implemented with several project partners, e.g. the Fraunhofer Austria institute or SZTAKI from Hungary.

Currently, I am engaged in a project with the insurance company called Zurich. I am mainly concerned with the implementation and automatisation of workflow tasks for mathematical models of catastrophic events (e.g. floodings, earthquakes, or terrorism). The programming languages used are mainly R, for which I have established several basic and reusable packages (persistence layer, API implementations, workflow automation packages, etc.) in the department, and SQL using the Microsoft SQL Server. For planning we use Azure DevOps.

Programming Languages: Procedural (e.g. C, R, LaTeX), Object Oriented (e.g. Java, Python, C++, R see R6) and Functional (e.g. Haskell, OCaml). I have worked with all of the mentioned ones, though the list is incomplete. See also my CV and my slightly outdated public Github profile.

Clients and Project Partners: Engel Austria, Fraunhofer Austria, SZTAKI (Hungary), Universität Innsbruck, VRVis, Robert Bosch AG, Zurich Insurance Group

Main Tools: Linux, Windows (98, XP, Vista, 7, 8, 10), Databases/SQL (MySQL, Postgres, MS SQL Server), Emacs, LaTeX, MS Office/Libreoffice, Concurrent Programming, Statistics, Machine Learning, Artificial Intelligence (AI), Manufacturing and Resource Planning (MRP), Supply Chain Management, Hierarchical Production Planning (HPP), Optimization (MILP, GRASP, ML and other heuristics), Simulations, Data-Warehouse Systems, Extraction-Transformation-Loading Process.

Natural Languages: German (native), English (business-fluent), Spanish (working proficiency)

Main areas of research knowledge: Operations Research, Production Planning, Order Release, Machine Learning, Simulation, Reinforcement Learning, Optimization, Statistics, Runtime Complexity Analysis.

You want to more info or are interested in working with me? I am officially self-employed. Send me an email to ms@globalpont.com or call me at +34611093771.

Location: I am located in Valencia, Spain. Hence, I offer my services remotely with about one weekly on-site visit per month. My main work areas so far have been Austria, Germany, Switzerland and Spain, but I am open to travel.

References

  1. Haeussler, Stefan, Manuel Schneckenreither, and Christoph Gerhold. “Adaptive order release planning with dynamic lead times.” IFAC-PapersOnLine 52.13 (2019): 1890-1895.
  2. Moser, Georg, and Manuel Schneckenreither. “Automated amortised resource analysis for term rewrite systems.” International Symposium on Functional and Logic Programming. Springer, Cham, 2018.
  3. Moser, Georg, and Manuel Schneckenreither. “Automated amortised resource analysis for term rewrite systems.” Science of Computer Programming 185 (2020): 102306.
  4. Schneckenreither, Manuel, Stefan Haeussler, and Juanjo Peiró. “Average reward adjusted deep reinforcement learning for order release planning in manufacturing.” Knowledge-Based Systems 247 (2022): 108765.
  5. Schneckenreither, Manuel, and Georg Moser. “Average Reward Adjusted Discounted Reinforcement Learning.” Adaptive and Learning Agents (ALA) 2022
  6. Schneckenreither, Manuel. “Dynamic Strategies for TCT.” Logic and Computational Complexity 2020
  7. Schneckenreither, Manuel, Stefan Haeussler, and Christoph Gerhold. “Order release planning with predictive lead times: a machine learning approach.” International Journal of Production Research 59.11 (2021): 3285-3303.
  8. Schneckenreither, Manuel, and Stefan Haeussler. “Reinforcement learning methods for operations research applications: The order release problem.” International Conference on Machine Learning, Optimization, and Data Science. Springer, Cham, 2018.
  9. Schneckenreither, Manuel. “Smart Decision Support Tools for Operations Management with a Special Focus on Reinforcement Learning.” (2021). PhD Thesis
  10. Schneckenreither, Manuel, Sebastian Windmueller, and Stefan Haeussler. “Smart Short Term Capacity Planning: A Reinforcement Learning Approach.” IFIP International Conference on Advances in Production Management Systems. Springer, Cham, 2021.
  11. Haeussler, Stefan, et al. “The lead time updating trap: Analyzing human behavior in capacitated supply chains.” International Journal of Production Economics 234 (2021): 108034.