Development

Betting on Development

Description

Singlar has its methodology for the development of cyber-physical systems of Artificial Intelligence towards the optimization of quality and in particular for the optimization of production processes in CI 4.0, always with special attention to cybersecurity and business continuity, as well as in the development of customized solutions that ensure the proper functioning of the so-called Critical Infrastructures.

In its generic methodology, natural phases are followed for these developments:

PHASE 1. DEFINITION OF INITIAL OBJECTIVE

PHASE 2. EXPLORATORY ANALYSIS AND DATA COMPLETION

PHASE 3. DEFINITION OF THE FINAL OBJECTIVE AND THEORETICAL PROTOTYPING OF THE ALGORITHMS

PHASE 4. PROGRAMMING OF THE ALGORITHMS BASED ON THE MODEL(S)

PHASE 5. EVALUATION AND TESTING. GENERATION OF RESULTS AND DELIVERABLES

PHASE 1. DEFINITION OF INITIAL OBJECTIVE

Define the requirements pursued by the cyber-physical system.

PHASE 2. EXPLORATORY ANALYSIS AND DATA COMPLETION

Perform an analysis of the available data to assess its adequacy and validity by preparing the data sets in order to achieve the objective defined in phase 1.

PHASE 3. DEFINITION OF THE FINAL OBJECTIVE AND THEORETICAL PROTOTYPING OF THE ALGORITHMS

Based on the results obtained from the exploratory analysis of phase 2, the nature of the final target metric to be used to measure the quality of the models is established.

This phase, together with phase 1, aims to satisfy the strategic objective of linking the application of AI in process modeling based on the current state of the art and theoretical foundations.

PHASE 4. PROGRAMMING OF THE ALGORITHMS BASED ON THE MODEL(S)

Programming of the computational prototype of the proposed algorithms and its training with the available retrospective (or prospective) data from phase 2.

Singlar uses a pseudo SCRUM method, usual practice and part of SINGLAR’s expertise, so that software development methodology based on SCRUM is applied in the different internal sub-phases of the development, partially selectively feeding a product backlog and incorporating the development by sprints, all with a framework based on CMMI and SPICE best practices.

detail of the SCRUM methodology

PHASE 5. EVALUATION AND TESTING. GENERATION OF RESULTS AND DELIVERABLES

Validation of the algorithm when feeding the datasets, either on retrospective or prospective data.

Request a quote

    Data controller: SINGLAR (Singlar Innovación, S.L.) | Purpose: To manage the commercial/professional relationship | Legitimation: Consent of the interested party/Execution of a contract | Recipient: No data will be transferred to third parties, except with express authorization or legal obligation | Rights: Access, rectify and delete data, data portability, limitation or opposition to its processing, transparency and the right not to be subject to automated decisions, for more information access the privacy policy.