Design and development of a platform for workers’ security monitoring in Industry 4.0 scenarios

In 2019, 503.790 accidents in the industry sector have been recorded in Italy, 974 of which involved death. Among the causes of accidents there are environmental causes: accidents happen because of the working environment and the triggering factors can include poor lighting, ambient temperature, air pollution, sound pollution. More recently, indoor air quality, particularly when concerning CO2 and Particulate matter, emerged as one of the most reliable proxies indicating the probability of viral infection transmission. This thesis presents the “SALVO” project, designed and developed with ENEA (Italian National Agency for New Technologies and Sustainable Development) and STMicroelectronics, that using IoT and Cloud Computing, aims at creating a distributed environment monitoring platform to improve workers’ security in an Industry 4.0 context. Its core component is a multi-sensor wearable device able to measure temperature, VOC, air composition, inertial parameters, sound, and location (both indoor and outdoor). The application requires low power, low complexity, and low costs nodes, deployed on a scalable network, that can communicate over long distances and in different scenarios. These requirements are satisfied by emerging wireless communication technologies called low power wide area network (LPWAN), which provide long communication ranges (up to 40 km in rural zones and 5 km in urban environments), with support of the simultaneous connection of thousands of end devices to the infrastructure. The technology chosen for “SALVO” is LoRaWAN, through which the device sends, via radio frequencies in the unlicensed EU868 ISM band, two types of messages: a low priority telemetry message, used to construct a historical trend of the environmental conditions and a high priority alarm message, triggered when some given thresholds are exceeded and representing a dangerous situation for the operator. A custom coding is introduced which allows sending all the information the node can provide in a smaller packet with respect to the standard low power payload encoding (CayenneLLP), resulting in better battery performances and delivery rates in harsh environments. The radio messages sent by the node are demodulated and converted into an IP packet by a gateway and then forwarded to a Network Server which manages the whole LoRaWAN network. Finally, an application server displays the data collected in a dashboard addressed to the security managers under the form of maps, heatmaps, and tiles. The proposed project allows for fast intervention when needed and allows the collection of environmental parameters that in a future version of the platform can be used to perform some AI algorithms for predicting dangerous situations, possibly reducing injuries and saving lives.