European Commission logo
polski polski
CORDIS - Wyniki badań wspieranych przez UE
CORDIS

Advanced monitoring system for horse riding and equestrian activities based on wireless sensing and artificial intelligence

Final Report Summary - EQUISAFE (Advanced monitoring system for horse riding and equestrian activities based on wireless sensing andartificial intelligence)

Executive Summary:
The EQUISAFE project places the horse and its well-being at the center of focus. All systems are designed around the horse’s complex physiology and psychology. The goal of the EQUISAFE project is to develop a novel electronic system that automatically monitors horse activity and health parameters. The EQUISAFE software platform collects all data received from the sensors and data input from the user to manage an up-to-date database of horses and to take timely informed decisions whenever a health-threatening situation is detected. The system automatically detects rider falls and accurately pin points the accident locations to relief personnel.

The system is designed to be user-friendly and intuitive for non-expert users.

The main project objectives are:

i) To identify the technological and functional system specifications. To create case studies for system validation.
ii) To design and build communication networks using short-range and long-range wireless systems that connect sensors to the central computer running the EQUISAFE software.
iii) To develop a portable system that assists guides during expeditions.
iv) To design and build a wireless inertial measurement unit that monitors horse behavior and detects abnormal movements.
v) To design and develop three health sensors (perspiration, pulse and oximeter).
vi) To design and develop a wireless fall detection sensor.
vii) To develop an intelligent camera that non-intrusively monitors the posture of horses inside stalls.
viii) To build and train the expert knowledge design system.
ix) To design and develop the software platform and user interface.
x) To integrate all components into one system and test it.
xi) To validate EQUISAFE in real-life scenarios.
xii) To train SMEs.
xiii) To disseminate project results and prepare for commercialization.

Project Context and Objectives:
EQUISAFE is an advanced monitoring system for horse riding and equestrian activities based on wireless sensing and artificial intelligence. The system consists of multiple micro-systems that are integrated together to provide an intelligent and reliable systems. Nevertheless, each micro-system can work autonomously, satisfying different customer and market requirements.

The Main Board (Work Package 2) was developed and tested in a real life scenario. The Main Board is battery powered and is fixed to the horse saddle. The device contains a GPS sensor to acquire real-time and accurate global positioning (longitude and latitude) and two wireless technologies: (i) short-range wireless contains XBEE 2.4GHz to communicate with other Main Boards (Guided tour) and the Helmet Fall Detector and (ii) long-range wireless contains the SIM908 GSM/GPRS module for long-range communication using the existing mobile infrastructure. The SIM908 also contains the GPS module, hence offering a very cost-effective solution for remote tracking. Through the EQUISAFE software (developed in Work Package 7) the owner of the horses can easily query the location of a specific horse and the status of the rider (fall, riding etc). When the system detects a fall, the Main Board transmits an emergency message to the EQUISAFE software and an SMS to up to five registered users. The horse owner and emergency rescue team are alerted instantly with the fall event, location of the fall and the status of the rider after the fall (moving or lying on the floor indicating a possibility of a concussion or death). Since the Main Board remains fixed to the horse saddle, the owner can receive location updates of the horse if this flees away from the scene of accident.

An elaborate Fall Detection Algorithm was developed in Work Package 4. To detect falls accurately a fall chord mechanism (same used in water sports, such as jet skis and power boats) was added to complement the limitations of the Helmet Fall Detector (developed in Work Package 4). We have achieved a failsafe system that does not trigger false alarms (annoying horse owner) and does not overlook a true fall (protecting the rider at all times). The Helmet Fall Detector is also battery powered and connects with the Main Board mounted to the rider’s horse via the XBEE 2.4GHz short-range wireless transceiver. The Helmet Fall Detector can attach to any helmet (shape and size), by means of an elastic mount making it universal. The Helmet Fall Detector was tested in the laboratory, using the EQUICHUTE simulator (twice) and real falls performed from a stunt man.

The Wireless Gateway was developed in Work Package 2. It is installed in the horse stables to collect all the data messages transmitted from the health sensors and Horse Inertial Sensor (also referred to as Horse Behaviour Sensor) and relay them to the PC running the EQUISAFE software (developed in Work Package 6 and Work Package 7), for further data processing and decision making.

In Work Package 3, the Horse Inertial Sensor was developed and tested many times. Two versions were developed, the most recent one being more advanced and very compact that it could be installed under the jaw, hence enhancing the safety and comfort for the horse wearing the sensor. This sensor measures the 3D acceleration, performs higher level processing to extract the mean and entropy for each axis (x, y and z). The features are transmitted to the PC running the EQUISAFE software, via the Wireless Gateway, where an Artificial Neural Network classified the type of behaviour. The Neural Network was trained using real movement measurements and post-analysis from horses at Haras du Pin and foaling mares at Chamberet. The system is capable to identify a specific behaviour (eating, sleeping, lying down lateral or ventral, nervous, colic/foaling) with over 96% accuracy.

In Work Package 4, the health sensors were developed and tested on-site. The perspiration sensor detects a change in sweating behaviour in the subject horse. It was validated by IFCE at Haras du Pin stables. The pulse sensor measures the heart beat of the horse. It was validated in horse stables in Spain (Girona) and by IFCE at Haras du Pin stables. The oximeter sensor did not give successful and reliable results, despite the numerous trials, hardware/software changes and sensor positioning. The consortium agreed to abandon the commercialisation of the sensor. However, the generated Intellectual Property can be used for other markets in the health measurement industry and the technological SMEs are interested to exploit the results. The perspiration and pulse sensors transmit the data to the EQUISAFE system, via the Wireless Gateway, for further processing and sensor fusion. The wireless fall detection, horse-to-rider distance and sensor fusion all fall under the Helmet Fall Detector introduced above.

The internet-connected camera was developed in Work Package 5. The hardware and software to capture the video, process it and stream it to the user was developed. The image processing algorithm intended to automatically detect the posture of the horse inside the stable proved to be very challenging and non-feasible due to many parameter settings and challenges created by the surrounding environment. Due to these reasons, the consortium decided to abandon the development of the image processing algorithm and focus on the commercialisation of a camera that streams video/still images of the horse showing distress or abnormal activity (nervous, colic, foaling) detected from the wearable Horse Inertial Sensor. When the users receive an alarm that a specific horse is behaving abnormal, he/she can stream instant video or images of the horse to make a visual decision. This strategy was accepted more by the customers we interviewed.

The Expert Knowledge Modelling System was developed in Work Package 6. This intelligent software is configured using equine expert advice provided by IFCE. Given a combination of events and sensor measurements, the software is able to make a decision, such as identifying an abnormal behaviour.

The user-friendly EQUISAFE software, developed in Work Package 7, provides the Human Machine Interface between the EQUISAFE sensors and the Horse owner (Human User). The interface is very graphical, making it very easy to use. Apart from viewing the alerts, the EQUISAFE system empowers the users to be more organised, efficient and up-to-date.

Validation of the system commenced at Month 13 (January 2015), where the RTDs verified and validated the functionality and performance of the individual sensors, and later the EQUISAFE system after components were integrated together. Most of the validation activities took place at Haras du Pin stables (IFCE). Other places where validation took place was in horse stables in Spain.

Training – Training activities were organised at every meeting. The final hand over (training session) was given at Month 24 meeting, which was held in Malta.
Dissemination – Dissemination activities were required in order to ensure the propagation of the results of the project beyond the consortium to a wider audience. All partners were required to take an active role in discussing and deciding the content of the published material, ensuring that no confidential information is divulged. An Exploitation Manager was appointed by the consortium based on his track record and experience in the field, where his role was to coordinate all activities encompassing any transfer of knowledge outside the Consortium; however particular actions were to be executed by designated partners. Conferences, scientific papers and events were to be coordinated by IFCE who took advantage of the events which they were regularly invited to. Dissemination material was edited and elaborated in the following formats: leaflets and posters, video and website. Furthermore, articles were elaborated and published in various different trade magazines, and other relevant publications EU-wide, including associated countries. The RTD partners took the lead in this sub-task given that they have more experience in publishing research and technical articles. Mailing of information regarding the project to interested parties was also carried out toward the end of the project. Information and exhibition stands were set up and manned at various events that were attended. A promotional video was also produced, highlighting the process of the project work and the benefits of the final product. This video was uploaded on the EquiSafe project website, and also on YouTube. A series of prestigious conferences were attended by the partners, where the project objectives, benefits and non-confidential R&D results were presented to interested parties.

Project Results:
The EQUISAFE project generated 6 results that can be commercialised:

1) Main Board for tracking location and status of horse and rider, anywhere and everywhere.

The Main Board is battery powered and is fixed to the horse saddle. The device contains a GPS sensor to acquire real-time and accurate global positioning (longitude and latitude) and two wireless technologies: (i) short-range wireless contains XBEE 2.4GHz to communicate with other Main Boards (Guided tour) and the Helmet Fall Detector and (ii) long-range wireless contains the SIM908 GSM/GPRS module for long-range communication using the existing mobile infrastructure. The SIM908 also contains the GPS module, hence offering a very cost-effective solution for remote tracking. Through the EQUISAFE software the owner of the horses can easily query the location of a specific horse and the status of the rider (fall, riding etc). When the system detects a fall, the Main Board transmits an emergency message to the EQUISAFE software and an SMS to up to five registered users. The horse owner and emergency rescue team are alerted instantly with the fall event, location of the fall and the status of the rider after the fall (moving or lying on the floor indicating a possibility of a concussion or death). Since the Main Board remains fixed to the horse saddle, the owner can receive location updates of the horse if this flees away from the scene of accident.

2) Helmet Inertial Sensor for detecting falls and monitoring of rider’s condition after the fall.

An elaborate Fall Detection Algorithm was developed. To detect falls accurately a fall chord mechanism (same used in water sports, such as jet skis and power boats) was added to complement the limitations of the Helmet Fall Detector (developed in Work Package 4). We have achieved a failsafe system that does not trigger false alarms (annoying horse owner) and does not overlook a true fall (protecting the rider at all times). The Helmet Fall Detector is also battery powered and connects with the Main Board mounted to the rider’s horse via the XBEE 2.4GHz short-range wireless transceiver. The Helmet Fall Detector can attach to any helmet (shape and size), by means of an elastic mount making it universal. The Helmet Fall Detector was tested in the laboratory, using the EQUICHUTE simulator (twice) and real falls performed from a stunt man.

3) Horse Inertial Sensor for detecting behaviours that require immediate attention (nervous, stuck in box, colic and foaling) while horse is unattended inside the stable.

The Horse Inertial Sensor was developed and tested many times. Two versions were developed, the most recent one being more advanced and very compact that it could be installed under the jaw, hence enhancing the safety and comfort for the horse wearing the sensor. This sensor measures the 3D acceleration, performs higher level processing to extract the mean and entropy for each axis (x, y and z). The features are transmitted to the PC running the EQUISAFE software, via the Wireless Gateway, where an Artificial Neural Network classified the type of behaviour. The Neural Network was trained using real movement measurements and post-analysis from horses at Haras du Pin and foaling mares at Chamberet. The system is capable to identify a specific behaviour (eating, sleeping, lying down lateral or ventral, nervous, colic/foaling) with over 96% accuracy.

4) Two health sensors: Pulse sensor for monitoring heart rate and Perspiration sensor for monitoring perspiration with reference to an activity.

The health sensors were developed and tested on-site. The perspiration sensor detects a change in sweating behaviour in the subject horse. It was validated by IFCE at Haras du Pin stables. The pulse sensor measures the heart beat of the horse. It was validated in horse stables in Spain (Girona) and by IFCE at Haras du Pin stables. The perspiration and pulse sensors transmit the data to the EQUISAFE system, via the Wireless Gateway, for further processing and sensor fusion. The wireless fall detection, horse-to-rider distance and sensor fusion all fall under the Helmet Fall Detector introduced above.

5) Internet-connected camera that enables stable and horse owners to stream and watch real-time video from their internet-connected devices when an unwanted behaviour is detected by the Horse Inertial Sensor worn by the horse.

6) EQUISAFE software that manages all the above systems, includes valuable software features for equine management and presents the results to the user through its user-friendly Graphical User Interface. Embedded in the software is the Expert Engineering Model that has been modelled from expert equine advice received from IFCE.

Secondary project result:

7) The oximeter sensor did not give successful and reliable results, despite the numerous trials, hardware/software changes and sensor positioning. The consortium agreed to abandon the commercialisation of the sensor. However, the generated Intellectual Property can be used for other markets in the health or sports measurement industry and the technological SMEs are interested to exploit the results.

Potential Impact:
The final result is the EQUISAFE advanced monitoring system for horse riding and equestrian activities based on wireless sensing and artificial intelligence. It provides an IT Platform for adequate intelligent monitoring of leisure equestrian activities. This was achieved by improving the safety and management of equestrian activities by means of on-line information of horses anytime anywhere. The system is modular with various optional add-ons that allow the creation of a system tailored to end-user needs. Furthermore, statistical information is analysed and timely advice is provided to the riders to avoid potential hazardous circumstances. The two-year project focused on leisure equestrian sport, and multiple sensing capabilities were developed.

The development of EquiSafe will introduce to the target sector of horse-riding schools a new product which will have beneficial aspects both from an economic as well as a social point of view: the availability of this technology provides a safer environment during horse riding activities, offering peace of mind to the users and employees and increasing the expansion of social business activities such as tourism. EquiSafe will minimise accidents which would otherwise end up on the media, thus improving the image of equestrianism as a safer sport that can be practiced by people of all ages. This will also positively impact new trends in the ecological tourism sector. These new tourism trends, as demonstrated, are of an increasing social interest as a promising leisure activity.

Besides the direct benefits to end-user SMEs, this project has a much wider societal impact as well as an impact on animal welfare, mainly:

• Provide a safer environment and a more reliable service for horse riding activities: - EquiSafe can detect a rider’s fall and its’ severity, as well as the location of the runaway horse. If a horse rider falls from a horse during an expedition, the accelerometers on the rider’s helmet sense the fall and can sense the severity of the fall based on the rider’s movements. If a severe fall is detected and the rider has limited mobility, EquiSafe assists the rider by sending up to 5 SMSs to preselected numbers with the longitude-latitude location of the rider and the horse. The system will also send a notification to the server, and the riding school will receive a notification on the PC.

• Provide a reliable horse health monitoring: - EquiSafe is critical in identifying Colic or Foaling. Colic is a clinical sign that a horse has abdominal pain which can prove fatal without surgical intervention. Among domesticated horses, colic is the leading cause of premature death. The horse inertial sensor and/or the movement analysis from the video feed of the installed cameras will register that the horse is in distress, and the horse owner will receive a notification and can view the horse form his mobile device or a PC. This way the owner can understand immediately what type of assistance the horse requires. The same applies for a foaling mare. Foaling is a rapid process and difficulties may arise quickly. Hence it is essential that any signs of abnormality are detected immediately. However, it is not practical for the owner or the vet to stand by and wait for foaling to begin, and EquiSafe can send a notification to the owner as soon as the video analysis from the video feed and/or the inertial sensor senses a distressed mare.

• Generate growth of social business activities: - EquiSafe increases the efficiency of the upkeep of horses, reducing costs on hiring a vet by means of early diagnosis of specific diseases and therefore increasing the wellbeing of horses. Moreover, EquiSafe provides peace of mind to the horse-rider as well to the horse riding schools as both the health of the horse as well as that of the horse-rider is monitored and risks of injuries is minimised. These two factors will collectively contribute towards the potential to generate growth of horse-riding tourist attractions.

The benefits of such an application derive from the fact that EquiSafe will use cutting-edge sensor and wireless technology as well as Artificial Intelligence based software system. The use of EquiSafe allows:

- A decrease on the risks associated to equine activities;
- An increase on the efficiency on the management of the horse population, increasing their benefit and business capacity;
- A decrease on the maintenance costs with the use of EquiSafe.

Moreover, horse-riding activities are lately gaining more social relevance in the treatment of mental disorders, such as autism or depression. EquiSafe contributes to a better horse-riding experience for such patients and will avoid these sensible riders with any unpleasant or unexpected situations during therapeutic rides. In such cases, it is important to avoid unexpected incidents as these might have the opposite effect on the patient and can result in serious consequences. In these situations an old and calm horse would be the ideal mount to be used for such activities, however horse behaviour is complex and difficult to predict completely, thus EquiSafe will contribute with its advice on the current status of horse temperament based on AI predictive models.

The EquiSafe project will also contribute to the R&D in a sector where technology is generally relegated to “second position”, thus ensuring that further development and the safeguarding of the well-being of humans in proximity of animals, and of the animals themselves is maintained.

Dissemination work

Many of the prestigious events attended (such as the Spoga Horse Fair) signifies high communication impact as such conferences attract thousands of visitors from academia as well as from the industry. Digital media such as the website, the video as well as the pdf leaflet and poster had an important role for wider coverage as dissemination was not restricted to time and location and the exploitation board could provide information for all interested parties anytime. Most of the actions were targeted toward the relevant industry segments to facilitate pre-marketing activity.

By employing the foreseen dissemination channels selected for the designated target audience the consortium carried out 34 dissemination activities in total. This entails the preparation of informative dissemination material including on-line information sharing, representation on thematic conferences, fairs and exhibitions.
The project partners deemed it important to have a strong online presence and promote the project as much as possible, and in fact quite a few dissemination activities were carried out during the project timeline, including the project website at www.equisafe-project.com the project video, leaflet, poster, newsletters, press releases on CORDIS and other relevant platforms. The leaflet, poster and the newsletters are available to download from the project website. Both the poster and the leaflet are available in English, French, Spanish and Russian.

More than 20 articles were published on equestrian websites, as well as on important influential equestrian journals, mainly: The Equine Chronicle (published in English), Ecuestre – La revista del mundo del caballo (published in Spanish) and L’équimag (published in French). Additionally, a peer-reviewed scientific paper “Automatic activity classification of horses using wireless body mounted motion sensors” (ISBN: 978-1-943842-25-4) was published in the 11th International Equitation Science Conference at the University of British Columbia, Vancouver, Canada.

The consortium also intensified its participation in several equestrian events where the EquiSafe project and the system itself were disseminated to their target market, such as the 2014 World Equestrian Games, the EQUITA 2014 fair, the Cheval Passion fair, the Journée de la Recherche Equine, Menorca Technological Fair 2015 and the Spoga Horse International Fair for Equestrian Sports 2015.

Exploitation of results

The commercial route envisaged for the exploitation of the results will be carried out via consortium members. The sales and marketing strategy is based on a business plans as detailed in the final PUDF. The marketing objectives and target market segments have also been identified by the exploitation board. Prospective customers have been shortlisted on a mailing list, and were contacted towards the end of the project with the project results and with the pricing of the system. Market changes and the activities of competitors were continuously being monitored by all the SMEs in the exploitation board (Market Watch), whilst the RTDs took into account changing customer needs and developments in technology (Technology Watch). The SMEs were constantly identifying existing, potential and key customers.

It was agreed by the exploitation board that the most adequate exploitation route is that the SMEs are directly involved in the sales of the system, without an intermediary. It is considered that the technical SMEs (EDE and SENIN) will be more directly involved in the day-by-day exploitation than the end-user SMEs (SARL). The commercial route envisaged for the exploitation of the results will take place through the consortium members, but also through external business partners.

The roles of the SMEs in the exploitation are as follows:

• EDE and SENIN (technical SMEs) – Responsible for the manufacturing, integration, QA as well as focus on the marketing and distribution of the EquiSafe system. SENIN will become the technical representative for Western Europe and the Mediterranean region, meanwhile EDE will become the representative for East Europe. The respective geographical areas where EDE and SENIN will carry out such Commercial Use is based on the technical SME’s current commercial links
• SARL (end-user SME) – experienced owners of a vast equine facility. Acted as a “test-lab” for EquiSafe during validation, and will use the prototype for demonstration purposes to interested parties.

The commercial EquiSafe system is module-based with various optional add-ons, and will consist mainly of the:
- Main board with the monitoring software and expert system;
- Horse inertial sensor;
- Helmet sensor (fall detector);
- Perspiration sensor;
- Pulse sensor.

A Business Plan for a 5 year period (2016-2020) was set up. During the market research phase of the project it was evident that there were no market studies or statistics available that identified the spending trends of horse riders or riding schools, their inclination towards buying safety equipment or how much are they willing to spend. In this context the exploitation board investigated further this market research gap and conducted an independent market survey during the second period of the project where it was analysed whether horse riders and/or riding schools were willing to buy or rent horse-riding safety systems such as EquiSafe. This investigation helped the exploitation board to be in a better position to formulate the final business plan rather than rely only on other statistical data. This survey helped address the specific questions that the exploitation board had, and was the only way that a realistic business plan could be prepared.

Based on the market survey carried out by the consortium, it was determined that the EquiSafe system can either be sold or hired. The strategy for pricing the product took into consideration the fact that there doesn’t exist an alternative product that can be compared to EquiSafe. Therefore potential customers do not have a reference to determine what would constitute a fair price. Too low a price might leave potential buyers unimpressed whereas too high a price might discourage the buyer. Hence the exploitation board’s primary goal initially is to disseminate the product’s benefits to the potential buyers and communicate EquiSafe’s value. The set prices are ideal for market penetration, as the exploitation board wants to go for a pricing strategy that focuses on their main market segment of buyers who are price-sensitive but who value not only the price but also the product.

A distribution of profits share is in place and spread among the exploitation board. EDE and SENIN hold 40% each, whilst SARL has 20% shares form the profits of sales of EquiSafe. With this profit distribution, an average of 160% ROI is envisaged for EDE, 155% for SENIN and 142% for end-user SME SARL.
The forecasted revenues are expected to be around €3.76M over a 5-year period, resulting in an overall direct profit of €1.79M translating in a ROI of 1.65.

List of Websites:

EquiSafe distributors:

ELECTRONICS DESIGN LIMITED
Tuulemae 5,
Tallinn 11411,
ESTONIA
Tel: +372 605 3060
Fax: +372 605 30 61
Email: electronics@electronics.ee

SENSOR INTELLIGENSE S.L.
Calle Vazquez de Mella 31,
Madrid 28017,
SPAIN
Tel: +34661794571
Email: info@galtel.es