-
AutoExp: A multidisciplinary, multi-sensor framework to evaluate human activities in self-driving cars
Authors:
Carlos Crispim-Junior,
Romain Guesdon,
Christophe Jallais,
Florent Laroche,
Stephanie Souche-Le Corvec,
Laure Tougne Rodet
Abstract:
The adoption of self-driving cars will certainly revolutionize our lives, even though they may take more time to become fully autonomous than initially predicted. The first vehicles are already present in certain cities of the world, as part of experimental robot-taxi services. However, most existing studies focus on the navigation part of such vehicles. We currently miss methods, datasets, and st…
▽ More
The adoption of self-driving cars will certainly revolutionize our lives, even though they may take more time to become fully autonomous than initially predicted. The first vehicles are already present in certain cities of the world, as part of experimental robot-taxi services. However, most existing studies focus on the navigation part of such vehicles. We currently miss methods, datasets, and studies to assess the in-cabin human component of the adoption of such technology in real-world conditions. This paper proposes an experimental framework to study the activities of occupants of self-driving cars using a multidisciplinary approach (computer vision associated with human and social sciences), particularly non-driving related activities. The framework is composed of an experimentation scenario, and a data acquisition module. We seek firstly to capture real-world data about the usage of the vehicle in the nearest possible, real-world conditions, and secondly to create a dataset containing in-cabin human activities to foster the development and evaluation of computer vision algorithms. The acquisition module records multiple views of the front seats of the vehicle (Intel RGB-D and GoPro cameras); in addition to survey data about the internal states and attitudes of participants towards this type of vehicle before, during, and after the experimentation. We evaluated the proposed framework with the realization of real-world experimentation with 30 participants (1 hour each) to study the acceptance of SDCs of SAE level 4.
△ Less
Submitted 5 June, 2023;
originally announced June 2023.
-
Knowledge-based multi-level aggregation for decision aid in the machining industry
Authors:
Mathieu Ritou,
Farouk Belkadi,
Zakaria Yahouni,
Catherine Da Cunha,
Florent Laroche,
Benoit Furet
Abstract:
In the context of Industry 4.0, data management is a key point for decision aid approaches. Large amounts of manufacturing digital data are collected on the shop floor. Their analysis can then require a large amount of computing power. The Big Data issue can be solved by aggregation, generating smart and meaningful data. This paper presents a new knowledge-based multi-level aggregation strategy to…
▽ More
In the context of Industry 4.0, data management is a key point for decision aid approaches. Large amounts of manufacturing digital data are collected on the shop floor. Their analysis can then require a large amount of computing power. The Big Data issue can be solved by aggregation, generating smart and meaningful data. This paper presents a new knowledge-based multi-level aggregation strategy to support decision making. Manufacturing knowledge is used at each level to design the monitoring criteria or aggregation operators. The proposed approach has been implemented as a demonstrator and successfully applied to a real machining database from the aeronautic industry. Decision Making; Machining; Knowledge based system
△ Less
Submitted 14 May, 2019;
originally announced May 2019.
-
Analysis of a site's integrity by 3D models and Integrated database, case study : the pic-du-midi high-mountain observatory (France)
Authors:
Michel Cotte,
Florent Laroche,
Matthieu Quantin,
Loic Jeanson,
Nicolas Bourgeois
Abstract:
The concept of "integrity", as currently used in the analysis of World Heritage sites or cultural landscapes mainly consists of 1) the composition of a given site, regarding its origins and its current state of conservation; 2) the visual and functional relationships between its components (attributes). One of the major questions is "what defines the origin period?". The integrity analysis has to…
▽ More
The concept of "integrity", as currently used in the analysis of World Heritage sites or cultural landscapes mainly consists of 1) the composition of a given site, regarding its origins and its current state of conservation; 2) the visual and functional relationships between its components (attributes). One of the major questions is "what defines the origin period?". The integrity analysis has to clearly understand and evaluate which tangible components exist, in order to correctly identify their origin period, estimate how much of the original structure and function remain. An additional difficulty rises in case of scientific and/or technical heritage assessment: the very historical necessity of updating and implementing technical/scientific innovations. These follow advances in science and/or technique(s), and lead to frequent successive changes, impacting the site's structure. Hence, for living sites, one cannot see the origin of the project as the only reference state, as it could be for classical heritage; it is instead requisite to enlighten a chronological series of major reference states. The Pic-du-Midi Observatory (France) is a remarkable example as high-mountain scientific station (2860 m). Its various and successive scientific uses led to a series of resha** and structural evolutions over around 150 years, till nowadays. We have there a series of reference states very conform to an active scientific station and observatory. The project aims to combine historical data and measurements on the current site's state, to build a succession of 3D models for a series of 5 or 6 'reference state' of the observatory within associated database, and to document the integrity analysis more accurately than usual. The final uses of the digital achievement will also serve for the management of the site and for further development of
△ Less
Submitted 15 January, 2019;
originally announced January 2019.
-
Process Information Model for Sheet Metal Operations
Authors:
Ravi Kumar Gupta,
Pothala Sreenu,
Alain Bernard,
Florent Laroche
Abstract:
The paper extracts the process parameters from a sheet metal part model (B-Rep). These process parameters can be used in sheet metal manufacturing to control the manufacturing operations. By extracting these process parameters required for manufacturing, CAM program can be generated automatically using the part model and resource information. A Product model is generated in modeling software and c…
▽ More
The paper extracts the process parameters from a sheet metal part model (B-Rep). These process parameters can be used in sheet metal manufacturing to control the manufacturing operations. By extracting these process parameters required for manufacturing, CAM program can be generated automatically using the part model and resource information. A Product model is generated in modeling software and converted into STEP file which is used for extracting B-Rep which interned is used to classify and extract feature by using sheet metal feature recognition module. The feature edges are classified as CEEs, IEEs, CIEs and IIEs based on topological properties. Database is created for material properties of the sheet metal and machine tools required to manufacture features in a part model. The extracted feature, feature's edge information and resource information are then used to compute process parameters and values required to control manufacturing operations. The extracted feature, feature's edge information, resource information and process parameters are the integral components of the proposed process information model for sheet metal operations.
△ Less
Submitted 9 May, 2016;
originally announced May 2016.
