This SARATROF project (Analytical System for the Aggregate Performance of Workers and Robots in Manufacturing Operations) is a technological innovation initiative of Gaiastech, the innovation space of the Xunta de Galicia. It is also part of the Conecta Hubs 2021 program, funded by the Axencia Galega de Innovación (GAIN) and has the financial support of the European Union (co-financed by FEDER funds under the Feder Galicia 2014-2020 operational program). It also receives support from the Second Vice-Presidency and the Department of Economy, Business and Innovation of the Xunta de Galicia.
Since 2021, Situm has been working with Ancora, Ledisson, Ednon and Beta Implants to create an analysis tool focused on analyzing the performance of factories. This tool considers the behaviour of machines, IT systems, workers, and robots to increase the efficiency of industrial production.
Year 1: obtaining geospatial information
During the first year, Situm’s work focused on sensory fusion and artificial intelligence to obtain geospatial information. Within the framework of this project, Situm obtains traceability indicators of the position of users and assets in the industrial environment. These indicators are processed and sent to the data aggregation and analysis meta-platform developed by the consortium for cross-analysis and extraction of complete information about the worker.
In addition, this year we visited the facilities of the end-user (Beta), analyzed the different workflows and movements, and deployed the Situm location system in these facilities for data collection. Finally, we developed web panels to consult the analytics of the number of visits per day/hour received by an area within an industrial building (developed in collaboration with Dimensiona during the project).
In collaboration with Prefapp, we adapted Situm’s cloud architecture to send the data obtained in real-time from the project to the common meta-platform and make them available through REST APIs that allow the integration of cartographic and geo-analytical elements. Finally, in collaboration with CITIC, work began on a tool to analyse the sensitivity of signal models to identify areas of malfunctioning of the geolocation system in this type of environment.
Year 2: geospatial positioning and analytics
During 2022, Situm technology (implemented during the first year) has been used to capture the positioning of the people involved in the manufacturing processes at the end user’s plant (Beta). A significant volume of data is generated that will later be used for extracting relevant geo-analytics (e.g., to analyze the traceability of production processes) and for integration with the rest of the subsystems. The architecture developed by Situm in collaboration with Prefapp is used and improved.
In collaboration with CITIC, the signal analysis tool whose development began in the first year is evolving, and we are analysing the data captured during the first year. In addition, we are refining this metric and tool, as well as the positioning system, based on the analyses carried out. Using Big Data techniques, it has been possible to detect all the positioning sessions with stability problems (out of all the data generated) and thus obtain an overall percentage of the number of users whose positioning has stability problems.
In addition, work is starting on the new functionality of the same tool that will detect positioning errors due to interferences, sensitivity of signal models, or Wifi network deployment problems, even before they occur. To this end, the Situm calibration process is rethought, creating a prototype that allows us to validate the calibration while it is being carried out before deploying the productive system.
This year, we also extended the capabilities of geo-fence-based analytics with the help of Dimensiona. These analytics, especially in combination with indoor location systems, are used for a multitude of use cases. In the case of industrial environments, we have found that there are several use cases with common analytics needs that we have developed in the second year.
One of the most common use cases in the industry is to add the ability to generate and evaluate work standards in manufacturing. That is, to generate any measurement (e.g., time spent in a manufacturing phase, mobility between zones, number of manufacturing area changes, etc.) and, with it, to generate and evaluate work standards:
- Generation of standards based on the measurement of residence and travel times during their tasks.
- Once the standard is generated, we apply a measurement of occupancy and mobility against the standard to identify problems at specific points in time.
- This can be done as many times as you want to measure changes applied to the production process.
The basic analytics that have been developed in this second year, which serve to generate standards and measure the impact of improvements in production processes, are as follows:
- The detailed time of each stay in a geofence. This year, we have added the ability to view in detail each of the visits to a geofence, allowing the system operator to know who has made the visit, the times of entry and exit and also the time that the operator has stayed inside the geofence.
- Time spent per employee in each geofence. During this second year, a second geo-analysis is generated (in collaboration with Dimensiona) consisting of crossing the data of visits to each geofence with the total time spent by each operator in each one of them.
- Traceability of an operator’s flow. This allows, briefly, to evaluate the traceability of the operator through the different areas and to what extent they are carrying out work in them (high accumulated stay times) or not. In case you want to analyze in more detail the traceability of each operator, a new analytic is generated that allows you to determine the time and visits of each of them to each geofence. It also includes a detailed table of each visit and time spent, as well as a percentage calculation.
- The number of stays in a geofence per hour. Additionally, we have increased the granularity of the report created during the first year, making it possible to break down the visits by hours (and cross-referencing this data with the stay times per geofence).
For the remainder of the project, Situm will continue to improve the geolocation data generated, considering the difficulties posed by industrial environments and generating geolocation-based KPIs that can be integrated into the meta-platform to obtain performance improvements in the factory.