WO2020025212A1

WO2020025212A1 - Device and method for detecting unsafe actions or unsafe situations during construction or repair of building structures

Info

Publication number: WO2020025212A1
Application number: PCT/EP2019/065502
Authority: WO
Inventors: Erik BRANTS; Sofie DE DECKERE
Original assignee: Renotec
Priority date: 2018-08-03
Filing date: 2019-06-13
Publication date: 2020-02-06
Also published as: BE1026510A1; FR3084766A1; NL2023254B1; BE1026510B1; DE102019004473A1

Abstract

Device (12) for detecting unsafe actions or unsafe situations during the construction or repair of building structures, characterized in that the device comprises at least one helmet (1) and comprises a computer system (13) located remotely from the helmet, the helmet (1) being provided with a camera (3) to take pictures of the environment, the helmet (1) and the computer system (13) being adapted for wireless communication with each other in order to allow the transmission of images recorded by the camera (3) to the computer system (13), the computer system (13) being adapted to perform image analysis on images received from the camera (3) and to recognize unsafe situations or unsafe actions.

Description

DEVICE AN D M ETHOD FOR DETECTING U NSAFE ACTIONS OR UNSAFE SITUATIONS DURING CONSTRUCTION OR REPAI R OF BUI LDI NG STRUCTURES

The present invention relates to a device and a method for detecting unsafe actions or unsafe situations during the construction or repair of building structures.

It is well known that the construction and repair of building structures entails safety risks for the people that execute these and for others that are present at the location in question. Due to legislation and other regulations as well as instructions from clients, the safety risk is largely limited. However, constant vigilance with regard to compliance with regulations and instructions is required.

This concerns in particular regulations with regard to the positioning and use of tools such as cranes and scaffolding, the correct storage of building materials, the presence of barriers that are used to protect against falling from a height, in particular barriers on scaffolding and barriers around holes and the correct use of personal protection equipment such as gloves, safety glasses, helmets etc.

However, it is not self-evident to be constantly alert to dangerous or potentially dangerous situations and there is a reluctance to address other people if such a thing is seen. Also, recognizing potentially dangerous situations and actions requires significant and repeated training.

The present invention aims to provide a solution for the aforementioned and other disadvantages and provides for this purpose a device for detecting unsafe actions or unsafe situations during the construction or repair of building structures, the device comprising at least one helmet and comprising a computer system located remotely from the helmet, the helmet being provided with a camera to take images of the environment, the helmet and the computer system being adapted for wireless communication to allow the transmission of images recorded by the camera to the computer system, the computer system being adapted to perform image analysis on images received from the camera and to recognize unsafe situations or unsafe actions. The building structures comprise, but are not limited to, all types of buildings, but also aboveground and underground infrastructure such as sewers, tunnels, lifts, cableways, bridges, canals, dikes, quays, support structures for pipes or cables and wind turbines.

The images may be still images or sequences of still images or moving images.

Hereby at least a part of those present at a construction site wears such a helmet, which takes pictures of the actions of others at the relevant construction site, but possibly also of the own actions of the wearers of the helmet and possibly also of the state of storage sites of building materials and tools for the construction and/or repair of building structures.

This has the advantage that more potentially dangerous actions or situations may be detected without the need for those present at a construction site to be distracted from their primary task, and in particular that not only people wearing the helmet, but also other people present, benefit from this.

This also prevents potentially unsafe situations or actions from not being detected due to loss of concentration of people that are traditionally already responsible for the detection of potential danger, such as construction supervisors.

A further advantage is that such a computer system may recognize a much wider range of potentially dangerous situations, in particular in the case of actions that are only seldom performed by those present on site, so that they may be unaware of the dangers of these actions.

The effect that is given to recognizing a potentially unsafe situation or unsafe behavior may vary depending on the need. This may be, for example: reviewing by a human expert, cataloging and making available to a responsible person or organization with delay for discussion, or immediately triggering an alarm. However, the consequences that are given to the detection of an unsafe situation or unsafe action fall outside the scope of the present invention.

Preferably, the camera is positioned such that in normal use of the helmet by a person, the field of view of the camera is located in front of the person.

In a preferred embodiment, the computer system is adapted to perform image analysis and to recognize unsafe situations or unsafe actions in that the computer system has access to a database with reference images of unsafe actions or unsafe situations and/or with reference images of safe actions or safe situations and is adapted to compare an image recorded by the camera with the reference images.

This has the advantage that the database with reference images may easily be adjusted if additional images of dangerous situations or safe situations become available. A specific database may also easily be used for specific groups of activities, such as underground activities.

Alternatively, the image analysis may be performed by a system with artificial intelligence.

The computer system is preferably adapted to be self-learning, wherein the identification of unsafe situations or unsafe actions is improved on the basis of feedback from a human expert.

In a preferred embodiment, the computer system is adapted to perform image analysis and to recognize unsafe situations or unsafe actions, in that the computer system is adapted to individually recognize devices and/or machines and/or tools on images received from the camera and in that the computer system has access to data indicating whether a device and/or machine and/or tool may be used.

This is especially useful for identifying whether there are devices and/or machines and/or tools present on a location where construction or repair of building constructions is taking place that should not be used there, for example on the basis of legislation, or because a valid inspection or test certificate is missing.

Examples are combustion engines that may not be used in enclosed spaces, or scaffolding that has not been inspected.

The machines may be individually recognized on the basis of a unique number that they already have, for example a machine number that is on a machine, or on the basis of visual characteristics specifically applied for this purpose.

In yet another preferred embodiment, the computer system is adapted to perform image analysis and to recognize unsafe situations or unsafe actions, in that the computer system is adapted to recognize persons on images received from the camera on the basis of facial recognition, and in that the computer system has access to data that indicate which persons may perform which actions and/or data that indicate which persons may be present in which places.

To this end, the computer system must, of course, have access to a database with photos of the face of the people that may be present.

As a result, for example, the presence of unauthorized persons may easily be noticed, as well as the performance of activities by persons who do not have the required qualifications for that type of activities.

In a preferred embodiment, the helmet is provided with a module to determine the geographic position of the helmet, the helmet and the computer system being adapted for wireless communication in order to transmit the geographic position of the helmet to the computer system.

As a result, the exact geographical position of the helmet wearer is known in case of an emergency. This may also be used to distinguish between activities at different geographical locations. After all, it is quite possible that a computer system is designed to receive and process communication from many different helmets, at many different locations.

In a preferred embodiment, the helmet is provided with a loudspeaker, the helmet and the computer system being adapted for wireless communication in order to control the loudspeaker by or via the computer system, i.e. to have sound produced by the loudspeaker.

This allows the computer system or a person to give instructions to the wearer of the helmet or to give an alarm signal.

In a preferred embodiment, the helmet is provided with a microphone, the helmet and the computer system being adapted for wireless communication in order to transmit a signal from the microphone to the computer system. This has the advantage that the user of the helmet may communicate with third parties.

In a preferred embodiment, the helmet is provided with a motion sensor, the helmet and the computer system being adapted for wireless communication in order to transmit data measured by the motion sensor to the computer system. This has the advantage that inactivity, for example due to an accident, of the wearer may be detected by the computer system.

The helmet is preferably provided with a light source to illuminate the recording field of the camera so that it may also work in low-light situations.

In a preferred embodiment, the helmet is provided with one or more gas sensors to measure the presence of toxic, flammable or explosive gases in the ambient air and/or to measure the oxygen content in the ambient air, the helmet and the computer system being adapted for wireless communication in order to transmit data measured by the one or more gas sensors to the computer system.

In certain circumstances, for example during renovations of sewers or when constructing or repairing buildings for the chemical or petrochemical industry, there is a risk of a too low oxygen content or the presence of toxic, flammable or explosive gases. In these preferred embodiments the device according to the invention may detect such a danger.

In a preferred embodiment, the helmet is provided with a fall sensor, the helmet and the computer system being adapted for wireless communication in order to transmit data measured by the fall sensor to the computer system, so that a fall of the wearer may be detected and the necessary action may be taken.

The invention also comprises a method for detecting unsafe actions or unsafe situations during the construction or repair of building structures, wherein use is made of a device according to the invention.

In a preferred variant, a person that is present at a location where construction or repair of building structures takes place, wears the helmet, whereby images of one or more other persons are made with the aid of the camera and these images are analyzed by the computer system in order to detect an unsafe action by one or more other people or an unsafe situation for one or more other people.

As a result, by means of wearing such a helmet by a limited number of people at the relevant location, a higher safety level is achieved for all those present at the relevant location.

The wearers of the helmet are then preferably people that may observe many others during their normal activities, such as supervisors.

In preferred embodiments of both the device according to the invention and the method according to the invention, the said unsafe actions or unsafe situations are: the non-use of personal protection equipment by a person and/or the absence of a barrier to prevent falling in places where there is a risk of falling from a height, such as around a hole or on a scaffolding or a floor of a building under construction. To clarify the invention, a preferred embodiment is described below with reference to the following figures, wherein

figure 1 shows a helmet in cross section that forms part of a device according to the invention; and

figure 2 shows the device according to the invention and the use thereof.

The helmet shown in the figure is an industrial helmet 1 that meets the requirements of EN397.

The helmet 1 is provided with an outer shell 2 and a plastic interior, not depicted.

At the front of the helmet 1 a camera 3 is positioned, as well as an LED lighting 4. The camera 3 is incorporated in the visor 5 of the helmet 1 and has a resolution of at least 1920 * 1080 pixels.

A microphone 6 and a loudspeaker 7 are also positioned in the outer shell 2.

The helmet 1 is furthermore provided, on the inside of the outer shell 2, with a GPS module 8 and a motion sensor 9.

The camera 3, lighting 4, microphone 6, speaker 7, GPS module 8 and motion sensor 9 are connected to a central control unit 10.

Under the outer shell 2, at the back, a battery 11 is disposed to provide the other electrical and electronic components with power for at least ten hours.

Optionally, the helmet 1 may be provided with an oxygen sensor and/or sensors for measuring the concentration of explosive, flammable or toxic gases and/or with a fall sensor. These sensors, if present, are also connected to the control unit.

The control unit 10 is provided with a CPU and a memory. Integrated in the control unit are a WIFI functionality, a 4G data functionality and a Bluetooth functionality. The control unit 10 is provided with a holder for a SIM card to enable the 4G data functionality and with a USB connection for reading out the memory and charging the battery.

The device 12 shown in figure 2 also comprises, besides the helmet 1, a computer system 13. This computer system 13 is adapted to receive images from a number of helmets 1 as described above and to compare these images with a first database 14 with images of dangerous or potentially dangerous actions or situations, and to thereby detect a dangerous or potentially dangerous action or situation on a received image. The computer system 13 is for this purpose connected to such a first database 14 for the transfer of data.

The control unit 10 is adapted to connect to the computer system 13 via its 4G data functionality or via its WIFI functionality, via a secure connection, via the internet, or alternatively via a separate computer network. Via this connection, data may be exchanged in two directions between the control unit 10 and the computer system 13.

The helmet 1 is adapted in such a way that the camera 3 in principle always records video images and permanently transmits them to the computer system 13 and that no special action is required from the wearer of the helmet 1 for this to happen.

However, depending on the quality of said components, the charging status of the battery 11 and the quality of the connection, a series of still images may also be recorded, for example with an interval of 1 second or 1 minute.

If the connection is lost, the images may temporarily be stored in the memory of the control unit 10 and later, once the connection has been re-established, may still be sent to the computer system 13.

The computer system 13 also has access to a second database, not depicted in the figures, which contains photographs of the faces of the persons who are permitted to be present at the construction site. This second database also contains a list of places where these people may be present and which activities that they may perform. This second database also contains a list that indicates which machines, tools and equipment may be used on a part of the construction site or on the entire construction site.

The use of the device is as follows and illustrated in figure 2.

A number of persons 15, but not necessarily all persons, at a construction site wear said helmet 1. The camera 3 of this helmet 1 permanently records images which are received by the control unit 10 and are sent via a 4G connection to the computer system 13. These are images of mainly persons 16 other than the person wearing the helmet 1.

The computer system 13 compares the received images with images from the first database 14 and determines via an image comparison algorithm whether there is a potentially unsafe situation or a potentially unsafe action taking place. If this is the case, the computer system 13 may automatically communicate this to the wearer of the helmet 1, via the loudspeaker 7, so that the wearer 15 of the helmet 1 may take the necessary action.

The computer system 13 also determines, based on the photos present in the second database, which persons are visible on the received images, and whether these persons have permission to be present at that particular location and to perform the relevant activities.

The computer system 13 also determines from the received images which individual machines, devices and tools are present, and evaluates whether these machines, devices and tools comply with the applicable certification requirements and whether they may be used at the relevant location.

It is not necessary for the wearer 15 of the helmet 1 to deliberately search for potentially unsafe situations or take potentially unsafe actions or action so that such a situation or action is recorded. The wearer 5 may simply perform his normal activities. As an example, in figure 2 it is shown how a person 16 is performing grinding work without eye protection and also does not wear a safety helmet. The camera 3 records these images, independently of the attention of the wearer 15 of the helmet 1, after which the computer system 13 analyzes these images and alerts the wearer 15 of the helmet 1 so that they may address the person 16 concerned about not complying with the applicable safety regulations.

The computer system 13 also compares the face of person 16 with the faces of the persons who may be present at the relevant location, and gives a warning to the wearer 15 of the helmet 1 if this proves not to be the case.

If the wearer 15 of helmet 1 is sufficiently close to the grinding machine used by the person 16, the camera 3 may detect the type plate of the grinding machine, such that the computer system 13 may recognize the individual machine number and determine whether the grinding machine has a valid inspection certificate based on the data present in the second database.

Depending on the circumstances, a report may alternatively be made by the computer system 13 which is subsequently discussed with the person in question 16 or persons, with the aim of achieving a long-term behavioral improvement.

In serious cases, such as for example a fire, the computer system 13 may automatically notify the fire brigade on the basis of the images.

If necessary for a good image formation, the lighting 4 may be switched on. This may be done manually or automatically.

The location of the wearer 15 of the helmet 1, and whether or not he moves, are recorded by the GPS module 8 and the motion sensor 9, and are also transmitted by the control unit 10 to the computer system 13.

Claims

1. Device (12) for detecting unsafe actions or unsafe situations during the construction or repair of building structures, characterized in that the device comprises at least one helmet (1) and comprises a computer system (13) located remotely from the helmet, the helmet (1) being provided with a camera (3) to take pictures of the environment, the helmet (1) and the computer system (13) being adapted for wireless communication with each other in order to allow the transmission of images recorded by the camera (3) to the computer system (13), the computer system (13) being adapted to perform image analysis on images received from the camera (3) and to recognize unsafe situations or unsafe actions.

2. Device according to claim 1, characterized in that said unsafe actions or unsafe situations are: the non-use of personal protection equipment by a person and/or the absence of a barrier in places where there is a risk of falling from a height.

3. Device according to any one of the preceding claims, characterized in that the helmet is a helmet (1) that meets the requirements of standard EN397:2012.

4. Device according to any one of the preceding claims, characterized in that the computer system (13) is adapted to perform image analysis and to recognize unsafe situations or unsafe actions in that the computer system (13) has access to a database (14) with reference images of unsafe actions or unsafe situations and/or with reference images of safe actions or safe situations and is adapted to compare images recorded by the camera (3) with the reference images.

5. Device according to any one of the preceding claims, characterized in that the helmet (1) is provided with an electronic subsystem (8) for determining the geographical position of the helmet (1), wherein the helmet (1) and the computer system (13) are adapted for wireless communication in order to transmit the geographic position of the helmet (1) to the computer system (13).

6. Device according to any one of the preceding claims, characterized in that the helmet (1) is provided with a loudspeaker (7), the helmet (1) and the computer system (13) being adapted for wireless communication in order to control the loudspeaker (7) by or via the computer system (13).

7. Device according to any one of the preceding claims, characterized in that the helmet (1) is provided with a microphone (6), the helmet (1) and the computer system (13) being adapted for wireless communication in order to forward a signal from the microphone (6) to the computer system (13).

8. Device according to any one of the preceding claims, characterized in that the helmet (1) is provided with a motion sensor (9), the helmet (1) and the computer system (13) being adapted for wireless communication in order to transfer data measured by the motion sensor (9) to the computer system (13).

9. Device according to any one of the preceding claims, characterized in that the helmet (1) is provided with a light source (4) to illuminate the recording field of the camera (3).

10. Device according to any one of the preceding claims, characterized in that the helmet (1) is provided with one or more gas sensors for measuring the presence of toxic, flammable or explosive gases in the ambient air and/or for measuring the oxygen content in the ambient air, the helmet (1) and the computer system (13) being adapted for wireless communication in order to transmit data measured by the one or more gas sensors to the computer system (13).

11. Device according to any one of the preceding claims, characterized in that the helmet (1) is provided with a fall sensor, the helmet (1) and the computer system (13) being adapted for wireless communication in order to forward data measured by the fall sensor to the computer system (13).

12. Device according to any one of the preceding claims, characterized in that the computer system (13) is adapted to perform image analysis and to recognize unsafe situations or unsafe actions, in that the computer system (13) is adapted to individually recognize devices and/or machines and/or tools on images received from the camera and in that the computer system (13) has access to data indicating whether a device and/or machine and/or tool may be used.

13. Method for detecting unsafe actions or unsafe situations during the construction or repair of building structures, wherein use is made of a device (12) according to any one of the preceding claims.

14. Method according to claim 13, characterized in that said unsafe actions or unsafe situations are: non-use of personal protection equipment by a person and/or the absence of a barrier in places where there is a risk of falling from a height.

15. Method according to claim 13 or 14, wherein a person (15), that is present at a place where construction or repair of building structures takes place, wears the helmet (1) and wherein with the aid of the camera (3) images of one or more other persons (16) are made and these images are analyzed by the computer system (13) in order to detect an unsafe action of the one or more other persons or an unsafe situation for the one or more other persons.