MELSOFT MaiLab acts as a dedicated virtual AI data scientist, helping companies to overcome the challenges of missing budgets and non-existent data analysts or AI specialists and is empowering them to realise future-oriented manufacturing strategies. Quick to deploy and with minimal training required, the solution bases recommendations and actions on intelligence derived from both live and historic data, without requiring users to have any specialist expertise. In effect, the platform uses machine learning (ML), a subfield of AI, to automate data gathering across a variety of systems, predictive model creation, analysis and the mining of large volumes of data.
AN INTELLIGENT ASSISTANT FOR QUALITY-ORIENTED MANUFACTURING
Ease of use start right from the installation, as the platform is accessed in a browser-based environment that does not require any additional software. It can run on any industrial PC, including Mitsubishi Electric’s MELIPC edge-computing solution.
Once installed, the MELSOFT MaiLab features an intuitive user interface with clear web-based visualisations. To further assist users without extensive programming skills, step-by-step guidance is provided. Also, the software helps users to understand what the data are suggesting while supporting them throughout all phases of a data analysis project. This is achieved by having the datasets being processed and analysis models created based on end goals selected by operators. In particular, these activities within the AI data science tool use Mitsubishi Electric’s proven Maisart AI.
Connected to the manufacturing system, the MELSOFT MaiLab was developed to support a wide range of different application scenarios and can also be tailored to each individual setup.
More precisely, it can be used in off-line mode to feed existing empirical data to develop or refine suitable predictive models and customised using open Python scripts. The tool can then be used for real-time diagnostics, providing the data generated as the production line operates to the algorithms and returning insights on the status of the line, its performance and how it can be optimised. The platform can also offer additional information and functions to address the needs and requirements of various departments as well as subject matter experts (SMEs). In addition, the information being processed and produced is used to continuously increase the accuracy of its algorithms to enhance its outputs and help companies drive productivity over time, in line with futureproof continuous improvement strategies. Even more, flexible licensing schemes are also available to address the specific needs of a company. For example, these include yearly subscription models or one-off payment options.
