Landslides and floods monitoring system

The challenge at hand is the alarming increase in human suffering and economic losses due to flooding and landslides, primarily driven by climate-related factors. This global challenge has intensified over the past two decades, claiming a devastating toll of over 1.23 million lives and impacting 4 billion individuals, with economic losses nearing an astonishing USD 3 trillion globally. Rwanda, notably vulnerable, ranks 154th in the ND-GAIN index (2021) for climate vulnerability.

Annually, disasters exert a considerable toll on Rwanda, a landlocked nation highly susceptible to the seasonal variability of flooding and landslides. Approximately 40% of Rwanda's land area is exposed to these hazards. Compounding the issue is Rwanda's status as the second most densely populated country in Africa, with a population expected to double in the next ten years. The nation suffers the challenge of recurring flooding and landslides, intensified by climate change, geological processes, population density, and its unique rugged landscape.

On 2nd and 3rd of May 2023 alone, rainfall-induced landslides and floods caused significant damage in the North and Western parts of Rwanda, resulting in 135 fatalities, 106 injuries, the destruction of over 5,000 homes, and extensive damage to critical infrastructure, including schools and roads. 20,000 people were directly affected and more than 50 000 people indirectly affected. Direct losses from May 2nd and 3rd alone exceeded USD 300 million, bearing dire implications for Rwanda's environment and the socio-economic well-being of its people.

Critically, Rwanda currently has a significant gap in climate adaptation initiatives, has no early warning system in place, and lacks reliable monitoring systems for floods and landslides. Consequently, there is a lack of real-time, precise, and actionable information crucial for reducing risks posed by disasters. An investment in a reliable early warning system would reduce loss of lives and economic impacts of the disasters for a cost significantly lower than the scale of losses experienced in the past.

We propose our iMaster/DocuCam system, a landslide and flood monitoring system that integrates AI and automated optical surveillance for near-real time data.

iMaster/Docucam is a server based system with an on-site agent. iMaters/ DocuCam consists on an on-site camera, a server and a client application accessible through any browser via internet. The system is designed to seamlessly integrate into any environment, capturing real-time high-resolution images using PTZ- Camera and collecting weather data through a co-installed local weather station. Furthermore, all captured images and weather data are swiftly transferred to a secure server infrastructure, where they are readily accessible for evaluation and utilization.

We integrate the iMaster/DocuCam system with community engagement through Simulation Exercises for increased community preparedness and response to the disasters. Our solution provides important data for quick decision-making during emergencies, ultimately enhancing climate adaptation efforts.

Additionally, our open-source initiative enhance innovative data usage, improve system evaluations, and incorporate multidisciplinary considerations.

• Our system provides timely updates with near real-time images and weather data.
• Our system consists of a backup feature that addresses potential data transfer challenges arising from bandwidth limitations. In instances where data transfer to the server is impeded due to bandwidth issues, the system holds the data until the bandwidth is restored.
• We have a well-structured database and a web-based API, which enables easy and fast data retrieval.
• Our different Artificial intelligence/Computer Vision algorithms can detect early signs and patterns of landslides and floods, and aim to predict hazards in time, allowing for early community response.
• Our system can generate automated change reports to notify relevant stakeholders about changes or potential issues to ensure effective communication and a robust system. Those alerts will not only be generated promptly but will also be easily accessible. The information is formatted in PDF, ensuring compatibility across various devices such as laptops and mobile phones. Additionally, alerts can be disseminated through multiple channels, SMS, WhatsApp, and email, ensuring widespread and immediate communication to relevant people.
• Our system can establish a comprehensive data-driven framework by integrating diverse data sources, e.g. satellite imagery and soil properties etc., and can be integrated with existing systems. This integrative approach enables our system to collect a wide range of information allowing for an overall understanding of the region. Additionally, a python framework for software development in settings with bandwidth limitations allows people to work (access the data) without direct connection to the server.
• Our system is designed to operate in different geographical and environmental settings including powering by solar panels, ensuring its utility across various regions. This has been proven by our currently installed systems, where they work even in difficult weather conditions.

The iMaster/DocuCam system reaches the local community at risk in Rwanda where the system is installed. These individuals represent the immediate users who have direct benefits from the system's implementation.

Moreover, meetings with local leaders, as well as engagements with representatives from Ministry in charge of disaster management, have been instrumental. Discussions have highlighted not only the interest but also the importance of the iMaster/DocuCam system in strengthening disaster preparedness, response, and resilience activities at all levels of governance.

The success of the system is evident in its positive reception and acknowledgment by key stakeholders. The emphasis on disaster management and resilience activities at both the community and governmental levels underscores the significance of iMaster/DocuCam in addressing critical needs within the region.

To refine and enhance the system functionalities, we are committed to additional locations in parts of Rwanda vulnerable to landslides and floods.

• We already ran the first successful prototype in Rwanda, which also qualifies the team the best to deliver this solution.
• Our team comprises Rwandans, which gives us a better understanding of the needs of the local people and eases the interaction with the local community.
• The technology is proven in other applications like in power stations monitoring and has shown its capability over more than 10 years.
• We have diverse backgrounds and experiences.
• To emphasise our commitment to global collaboration, we have in our team an international expert in disaster preparedness. This facilitates the exchange of knowledge and best practice and ensures that our solution benefits from global expertise in this area.
• In addition, we actively promote collaboration with governments to share knowledge and strategies to further improve the effectiveness of climate resilience efforts.

The iMaster/DocuCam system for landslide monitoring is currently in the pilot stage. The iMaster/DocuCam system has successfully undergone an initial pilot in the Rambura sector, Nyabihu district, Rwanda, and actively monitored an area of approximately 12 square kilometres. Since September 2022 to-date, over one million optical images have been collected at a rate of 4000 images per day. Several millions of local weather data were also collected. We need to conduct an extended pilot in different landscapes of Rwanda and with varying geo-environmental conditions.

In May 2023, our system was the first to provide near-real-time evidence of rainfall-induced landslides in the area of coverage. Currently, our system is the only one continuously monitoring (before-during-after) landslides in Rwanda.  This data set is a gold mine for developing AI/Machine learning algorithms for detecting or predicting landslides and floods. These models are necessary to automate early warning systems for the hazards.

Coupling the iMaster/DocuCam system with Simulation Exercises encourages integration of the technology into the existing risk communication systems up to the local level. SIMEXs allow us to capture local knowledge and feedback while sharing expertise on the hazards, disaster preparedness, response, and climate adaptation in general.

The genuine interest and demand for iMaster/DocuCam are demonstrated through active engagement with users, meetings with local leaders, and interactions at national level. The system's importance in strengthening disaster preparedness, response, and resilience activities at all levels is highlighted through user validation and care, reflecting its positive reception and impact.

The iMaster/DocuCam system is performing effectively in both Germany and Rwanda, demonstrating its adaptability across diverse geographical conditions. This is exemplified through inclusive testing, ensuring the system's consistent functionality in various landscapes. The incorporation of a web-based graphical user interface accessible anywhere via the internet further underscores efforts to address diverse user needs and promote inclusivity in the deployment of this technology.

We are applying to Solve because we see its potential to provide invaluable support across various dimensions.  

• We need financial, legal, and mentorship support to foster our open-source initiative.
• We need financial support to enhance our system functionalities.  
• We need opportunities to connect and network with impressive peers.
• We need to gain cross-sector collaboration to overcome various barriers collectively.

• Our solution employs the innovative iMaster/DocuCam system, protected by patent EP3236440A1 from the European Patent Office.
• Our system sends automated change reports to local communities and authorities, with the aim to minimise casualties and property damage.
• Integration within existing communication channels ensures rapid dissemination of warnings, reinforcing overall emergency preparedness.
• Our scalable system adeptly monitors vast areas, adapting seamlessly to diverse environmental conditions. This capability enables the system to perform effectively across diverse landscapes and climates.
• The Geo- Location feature allows for integration with other data sets.
• Our User-friendly web-based interfaces facilitate easy data interpretation, aiding effective decision-making.
• Long-term storage for over 30 years.
• We are in the process of introducing an open data approach. By introducing open data policies, we aim to encourage collaboration with researchers, government agencies and universities. This strategy promotes a collaborative effort in landslide and flood risk management.  The openness of data facilitates collaboration and strengthens partnerships that contribute to effective and comprehensive solutions for disaster risk mitigation.
• Community engagement is an important aspect of our preparedness and response approach, which emphasises the active participation of the local community. This focus on community participation allows for a better understanding of the complexities associated with the  disasters.  This inclusive strategy not only increases the effectiveness of our initiatives, but also ensures that solutions are adapted to the  needs and circumstances of the local area.
• Operating with minimal services, our system ensures reliability and accessibility.
• Continuous improvement, guided by feedback from Simulations Exercises (SIMEX), propels us ahead of evolving landslide risks, maintaining the project's efficacy and relevance over time.

Inputs

• Resources to buy hardware ( PTZ Camera, Weatherstation, … and associated licences).
• Mobile data
• Travel
• Skilled personnel.
• Meeting rooms, materials, equipment.

Activities

• (0) Stakeholder engagement
• (1) New installations
  • Identifying new locations
  • Engineering of each location
  • Pre-installation and Installation
  • Configuration
• (2) Training
  • Workshops and sensitization
  • Simulation Exercises
• (3) Operation
  • Landslides and floods monitoring and data collection.
  • Improvement of evaluation for early warning system
• (4) Open-Data + Open-Source
  • Getting legal regulations
  • Provision of Open source and open data
  • Motivate the open-source contributors and researchers
  • Collaboration with universities
• (5) Inclusion of Geolocation in the system

Outputs

• (0) Key stakeholder engagement. MOUs and paid contracts signed.
• (1) Big coverage system from different areas of the country to ensure the provision of enough informative and diverse data.
• (2) Relevant information which helps in the enhancement of the system and helps in decision making.
• (3)(4) Improvement of early warning systems according to environmental impacts, sociological aspects and terrain.
• (5)Data interconnection for fast data availability and clear visualisation.

Outcomes

• Increased knowledge on disaster preparedness and response
• Increased data repository
• Capacity building
• Improving the land use planning
• Increased understanding of landslides and flooding.
• Senseful reaction of the trained people through Simulation Exercises towards disasters.
• Financially sustainable venture.

Impacts

• Ensuring safety of the local people.
• Reduced economic losses from landslides and floods.
• Empowering local people, businesses, and communities.
• Enhancement and acceleration of research through digitalization.
• Sustainable development

Impact metrics:

• Number of jobs created
• Number of installations and area covered
• Volume of data stored
• Number of Simulation Exercises and trained people
• Number of activities on GIT and number of new evaluation types
• Number of data users
• Comments and feedback showing interest
• Number of publications based on the data and number of universities working with the data.
• Number of changes to nationwide land use plans
• Population warned to move to safety and response time
• Number of emergency kits distributed
• Number of landslides and floods detected.
• Changes to land use plans in monitored areas attributable to project results.    

Measurability:

Number of jobs created: Quantity of employment opportunities in different fields (engineering, social sciences, …). This will have an impact on the economy of the local community and project.

Number of installations and covered area: This will provide data on population reached and ultimately impacted, hence the improved disaster resilience.

Volume of data stored: This will check the efficiency of monitoring across installations and in time.

Number of Simulation Exercises and trained people: This will indicate the level of disaster and risk awareness, helping in decision making.

Number of data users, Number of publications based on the data, and number of universities working with the data: This will indicate the ongoing need for information and indication of new evaluation types. It will also illustrate the level of engagement with local communities, researchers, and governmental bodies.

Number of activities on GIT, number of new evaluation types, and comments and feedback showing interest: This will indicate the interest and engagement with the datasets by the open-source community.

Population warned to move to safety and response time: This will indicate the effectiveness of our change reports. The response time will help to measure the success of the Simulation Exercises and workshops. The iMaster/DocuCams will be configured to monitor safety places to be able to calculate response times during emergencies.

Number of landslides and floods detected: This will indicate the need of our system in the region.

Number of emergency kits distributed: This will indicate success of collaboration with Ministry in charge of disaster management during Simulation Exercises.

Changes to land use plans in monitored areas attributable to project results: This indicates system abilities to detect landslides (both small and large scale) and the level of engagement with the local authorities in the areas.

The technology is about optical monitoring of landslides and foods. It is driven by experts from various fields of knowledge (Physics, IT, Geology, Remote Sensing, Disaster risk reduction) from three continents.  The work  spans from the acquisition of data images, recording, storage, evaluation, up to real world consequences of Simulation Exercises as recommended by the Sendai Framework.

• DocuCam is at the top of a hill, taking 4000 images each day over years of its landslide prone surrounding. This has been proven in the night of the disaster (2nd to 3rd May 2023) in Rwanda. Our system recorded the image before, during and after the disaster.
• Data are automatically transferred to a server (in Rwanda at Rwanda Space Agency (RSA)). RSA is responsible to provide all IT-services and information of earth related data in Rwanda to enable data access, reliable predictions, and responses.  As a Critical infrastructure, special technical and legal demands on access and security of the data are needed.
• The data at RSA are very versatile and are intended to be publicly available partly. There is a wide range of methods to evaluate the DocuCam images, not only geological ones. We attach great importance to the capacity building of local experts and local leaders for a better understanding and evaluations of the natural disasters. During the implementation we engage students to make them the future expert for the sustainability of the solution.
• We intend to create an open-source project to contribute to the development of various evaluations. Measures to prevent people to attack the critical infrastructure of server, where data are stored, will be taken.
• To generate impact of data and evaluation, the locals need to be prepared through Simulation Exercises according to Sendai framework.

full-time staff: 2

part-time staff: 3

contractors: 2

4 years

• Diverse Global Team: Our team includes men and women from Africa, Europe, and Asia and have equal rights to express their ideas and abilities.

• Inclusive Recruitment: Student consultants and experts are recruited based on merit, regardless of gender.

• Collaboration with District Offices: We collaborate with the district disaster risk management officer and partner with the social support unit for inclusive activities like SWOT Analysis and SIMEX engaging all genders.

• Community Sensitization at the Church.

• In accordance with the government, there is gender inclusion reflected in community leadership.

• Universal Benefits.

• Sensitization during communal work and workshops where everyone is invited.

• Everyone is invited to different meetings regardless of gender or background and every participant has the right to provide feedback.

• Public Talks:  we encourage equal participation and opportunities irrespective of gender. e.g: COP28 representation by woman.

• recruitment and payment strategy: work opportunities and equal payment are given to qualified people regardless of the gender.

• Open data: This ensures transparency and equal access to information.

Beside landslide and flood monitoring, we plan to expand use-cases to energy supply, mining and construction monitoring of buildings and roads for progress, quality control, and preemptive maintenance, diversifying revenues.

Public funding in the form of donations is required to cover costs at the pilot stage. By engaging stakeholders and demonstrating the impact of our solution on the local communities at risk, we will attract investment from both the public and private sector customers and generate sustainable revenues for operation as a start-up in Rwanda.

We are looking to partner with the International Federation of the Red Cross (IFRC) Rwanda during our simulation exercises and for their implementation of forecast-based action in Rwanda. IFRC’s current  focus in Rwanda (2023) includes improving the quality of its response to people’s needs – specifically, scaling-up the provision of cash and vouchers for dignified assistance in ways incorporating protection, gender and inclusion into cash programming.

Similarly, the Rwandan Red Cross is part of the IFRC Pan-Africa Zero Hunger Initiative which aims to strengthen food security and livelihoods through cash modalities as safety nets. This includes planning for multiple cash and voucher assistance payments (according to the seasonal calendar).

The iMaster/DocuCam system and the associated simulation exercises can support the Rwandan Red Cross operations through risk mapping, contingency planning, and feedback via the images and from the local communities during simulation exercises.

The initial pilot focused the iMaster/DocuCam system on landslide monitoring and was funded by GIZ through the Fair Forward Artificial Intelligence for All Program. We piloted our technology at a single site in Nyabihu District in Rwanda, covering approximately 12 square kilometers.

We are looking for further donor support to scale to other locations including multi-use of our solution for both landslides and floods monitoring. Doing this will, (1) increase the impact to people at risk in Rwanda, and (2) attract and convince a wider customer base.

We will use a dual-licensing model to serve our iMaster software to the Rwandan market. Our pricing model emphasises a balance in the principles of openness and community collaboration, with the need for sustainable funding as a private company.

The revenue sources and respective pricing models are:

• Sales of iMaster/DocuCams with licensed iMaster software, as a one-off payment for unlimited access, or on a subscription basis.
• Sales of servicing, support, and maintenance contracts. The contract terms vary by the terms of access to the hardware.
• Sales of additional features developed in the software as a one-off payment to become open source later.
• Sales of data provision contracts to private users including hosting, managing, and maintaining the iMaster software on behalf of the clients.
• Sales of training for simulation exercises.
• Sales of custom software development contracts.
• Paid collaboration partnerships to extend other solutions using the iMaster/DocuCam system.

We anticipate more revenues from service contracts than from the software itself.