Capture of the operational Traces of the company’s actors to build human Capital and define the winning Processes – CaTCaP

It is about proposing methods allowing to:
(1) observe and identify learning activities and situations: we propose a mixed approach characterized by the collection and analysis of traces resulting from human observation and activity analysis (video recordings, questionnaires, etc..) as well as operational traces existing in computer systems (log files, modeled traces, etc.).
(2) extract skills from the analysis of operational traces: once the traces are collected, we identify the relevant links between these traces and skills, identified individually and collectively as relevant to carry out the project activities at Energy Pool. Synergies between these skills are also considered.
(3) formalize and represent identified skills considering their nature (soft or hard) and dimension (individual or collective). Respect the personal data confidentiality is considered during this phase. We represent the skills as a mapping which evolves dynamically and cane b reconfigurable according to the exploitation’ objectives.
(4) identify and evaluate practices from the peers’ point of view: in parallel with observations, interviews with project’ members will make it possible to confront the observed people with the traces of their activities in order to reach the sense that they associate with their practices.
(5) identify and deploy key success factors during teams’ definition: we develop a correlation analysis between skills, resources, traces and practice’ evaluations in order to identify key factors that led to different performances.
These methods and mapping will be validated through two industrial cases and supported with a software demonstrator.

The work carried out in this project has led to the definition of methods for: 1) Observing learning situations, by collecting “informal activity traces” resulting from human observation and “organization traces” from computer systems, while respecting the legal and ethical conditions for data collection. 2) Identify skills based on the analysis of these traces. A mixed approach based on ethnographic analysis and semantic analysis (using NLP-type methods) was proposed. 3) Identify the key factors leading to performance, based on peer evaluation of the various practices capitalized in the traces collected. 4) Formalize and represent skills and effective practices in the form of graphs, responding to needs for expressiveness and conciseness, evolving dynamically and reconfigurable according to operating objectives. These methods break with the prescriptive and hierarchical approaches of traditional work and project management. A software demonstrator has been developed to tool these methods and facilitate their implementation on an industrial site. All the tools defined in the project (methods and demonstrator) were tested in the industrial field, considering various commissioning projects at CMDL (project partner).

The perspectives considered must take into account the aforementioned problems and the individual must regain the central place of the system and maintain control over it. The skills-based approach that we defend in this project leads us to place the individual at the heart of skills models where the individual drives the system and not the other way around. In our current work on skills, we are interested in an intermediate level of analysis, called the «meso level«, i.e. relating to organizations (e.g. companies, teams, etc.). New representations of skills based on our current models are being developed, and aim to bring this «meso« dimension to skills in order, for example, to provide organizations with strategic management tools. The prospects for developments relate both to the potential of artificial intelligence techniques, to extract information from data that would not be structured, and to the improvement of human-machine interfaces built via UX-Design to facilitate their use. These developments have been very beneficial for our abductive method, i.e. based on experiments to build a theoretical framework, before returning to new experiments. Other perspectives, in line with lots 1 and 2, are currently being developed. They concern, among other things, the exploitation of graph theory to support the interpretation of skills graphs for example. We are currently developing new visualization tools to try to respond to new issues related to skills.

Mlaouhi, K.; Cholez, C.; Gzara, L. An Action-based Model to Identify Human Competencies through the Trace of Actions: Case of a Building Energy Engineering Company. IFAC-PapersOnLine. 2022, Vol 55 (10), 169-174.
Bemmami, K.-E.; Gzara, L.; Maire, J.-L.; Courtin, C. From digital traces to competences. IFAC-PapersOnLine. 2022, Vol 55 (10), 1944-1949.
Mlaouhi, K.; Gzara, L.; Cholez, C. Use of Competency Management Methods and Tools in Project-based Organizations. Conference CIGI-Qualita, 5-7 mai 2021, Grenoble, France.

Eddin Bemmami, K.; Maire, J.-L.; Gzara, L.; Courtin, C.; Pouydebat, O. Toward A New Model Of Competences In Work Situations. INCOM 2021, Jun 2022, Budapest, Hungary. 1150 – 1155.

In an industrial context characterized by a high complexity in business processes and proliferation of information systems, the company's actors implement many learnings and practices to overcome this complexity. These learnings and practices encompass implicit and explicit competences that we aim to extract, manage and connect to business processes and products so to enhance the company’s skills and capitalize their reuse.

The objective of this project is to extract the skills encompassed in current practices and learnings in order to manage the intellectual capital of a company by developing and / or by changing the existing skills, and Intelligently mobilize the actors’ skills by defining suitable practices to real skills.

It is about proposing methods allowing to:

(1) observe and identify learning activities and situations: we propose a mixed approach characterized by the collection and analysis of traces resulting from human observation and activity analysis (video recordings, questionnaires, etc..) as well as operational traces existing in computer systems (log files, modeled traces, etc.).

(2) extract skills from the analysis of operational traces: once the traces are collected, we identify the relevant links between these traces and skills, identified individually and collectively as relevant to carry out the project activities at Energy Pool. Synergies between these skills are also considered.

(3) formalize and represent identified skills considering their nature (soft or hard) and dimension (individual or collective). Respect the personal data confidentiality is considered during this phase. We represent the skills as a mapping which evolves dynamically and cane b reconfigurable according to the exploitation’ objectives.

(4) identify and evaluate practices from the peers’ point of view: in parallel with observations, interviews with project’ members will make it possible to confront the observed people with the traces of their activities in order to reach the sense that they associate with their practices.

(5) identify and deploy key success factors during teams’ definition: we develop a correlation analysis between skills, resources, traces and practice’ evaluations in order to identify key factors that led to different performances.

These methods and mapping will be validated through two industrial cases and supported with a software demonstrator.