Drones for Mobility solutions as all-in-one service

AI Quick Summary

MobiLysis provides mobility solutions as all-in-one service based on drone dataThrough this challenge, we aim to start developing an online service delivered as software as a service (SaaS) to our clients and provide added value to our current business model. This platform will empower our clients to visualize the collected data, access results, and glean insights seamlessly. It is a proactive step towards enhancing user experience and facilitating informed decision-making based on real-time analytics. ProblemCity mobility is a complex dynamical system with different transport modes that interact and compete for limited space. This dynamic environment directly impacts economic sustainability, environmental concerns, and contributes to delays caused by inadequate infrastructure, high levels of motorization, and ineffective traffic management. The increasing complexity of mobility needs and services necessitates the traffic monitoring and the analysis of multi-sensor data, which traditionally consumes considerable budget and time resources and presents challenges related to data formats, quality assurance, privacy, central integration, and compatibility.Modern cities are equipped with car-centric equipment that cannot adapt to address their evolving mobility needs. Our solution addresses these pressing challenges by harnessing the power of drone swarms to provide high-quality data and fact-based solutions. By leveraging cutting-edge technology, we offer comprehensive insights and actionable recommendations, enabling businesses and municipalities to navigate the complexities of urban mobility with confidence and efficiency. Why Drones?Our vision is to reshape urban mobility by identifying sources of traffic congestion and mobility issues and provide actionable solutions. However, current methods rely on outdated and inefficient equipment, leading to unreliable data and a lack of comprehensive insights into mobility patterns across different locations, times, and transportation modes. This results in insufficient data quality to effectively address congestion issues. Common practices such as manual counting, loop detectors, and CCTV systems are prevalent due to their simplicity, but they are labor-intensive and rely heavily on manual effort.Moreover, these methods fall short in understanding the broader traffic context and identifying the root causes of congestion. Thus, they fail to provide a complete understanding of traffic dynamics and hinder the development of effective solutions.Additionally, most of the above are not designed for "soft modes" (bicycles, scooters etc.) or pedestrians even though they are crucial for sustainability. Monitoring mobility from the skies is an ideal solution due to latest advances in interdisciplinary topics; the huge advances in drones’ materials and technologies and secondarily advances in Computer Vision, AI and computational power. This is a strong alternative to traditional monitoring techniques as different transport modes (even soft modes) can be identified.Our solutionWe mix a blend of interdisciplinary expertise ranging from traffic engineering to route optimisation, machine learning and automatic control, through computational tools that enable deployment worldwide.Our aim is to revolutionise how emerging technologies like drones can accurately monitor the various mobility interactions and provide fact-based solutions to reduce their effect on traffic and the environment. While the use of drones for mobility purposes has previously been discussed, challenges associated with the technology had kept its adaptation in low levels, with the examined use cases focusing on small parts of a road arterial or 1-2 intersections.Since our incubation period at EPFL, our team has been actively exploring opportunities to integrate drones into existing smart city infrastructure. Our unique experiments with drone swarms have paved the way for valuable contributions to urban planning and traffic management.Using our optimisation algorithms and by taking into account the drone flight regulations, our team designs the drone flight plans which are executed by professional certified pilots with commercial drones for enhanced safety and standardisation. With their birds-eye view the drones are collecting videos in a privacy friendly way, as no vehicle plates nor faces can be identified. The collected videos are then being analysed using an already trained computer vision algorithm which is then fined tuned based on each city’s needs.We transform the drone into a multi-sensor for sustainable citiesOur vision is to provide fact-based solutions to cities that aim to become more sustainable. Thus, access to emission data is crucial. Using state-of-the-art models and the high-quality data that are produced through our method we can transform the drone into a multi-sensor that can provide strategic CO2 emissions or noise emissions maps in low-cost, non intrusive way. The integration of drones with cutting-edge estimation models removes the necessity for additional sensors either on the drone itself or within the road infrastructure. Hence, our solution can not only analyze and detect traffic congestion and mobility issues but also provide CO2 and noise emissions data for strategic decision makring for urban planners and policy makers.Our teamOur team is composed of 5 enthusiastic people with experience in cross-disciplinary topics ranging from traffic engineering and computer vision to optimization, control and data science. All of them are closely associated with EPFL. The members of our team are the following:Dr. Manos Barmpounakis, CEO and co-founder (EPFL Postdoc)Prof. Nikolas Geroliminis, Scientific Advisor and co-founder (EPFL Full Professor)Dr. Taylor Mordan, Senior Computer Vision Engineer (EPFL PhD and EPFL Postdoc)Dr. Dimitris Tsitsokas, Senior Data Scientist and Transportation Engineer(EPFL PhD and EPFL Postdoc)Mr. Jasso Espadaler Clapés, Senior Data Scientist and Transportation Engineer (EPFL MSc – Best average grade 2021 class) Professor and co-founder Nikolas Geroliminis is a world renowned researcher with more than 15 years of experience in urban traffic networks and currently serves as Editor-In-Chief of Transportation Research part C: Emerging Technologies journal. Thanks to our EPFL background, the team accounts for numerous scientific publications in top-quality scientific journals. In addition, our advisory board consists of Simon Johnson (mentor and investor in Swiss tech start-ups and advisor to the Swiss National Centre for Robotics) and Dr. Frédéric Dreyer (mentor and coach for start-ups and SME, accredited by Innosuisse and Cantonal Platinn institutions). Finally we should note that our EPFL spin-off is part of the “Initial Coaching supported by Innosuisse”.Future StepsOur team is dedicated to advancing the technology we have developed, currently at TRL levels 7-8, by integrating data from additional sensors such as loop detectors and cameras. By continuously assessing the accuracy of the data collected and estimated, we envision to fast-track the integration of drones into Intelligent Transportation Systems (ITS) infrastructure. Our goal is to offer a cost-effective, efficient, and privacy-conscious multi-sensor solution. Leveraging this technology, we're already making strides in identifying interactions between "soft modes" and pedestrians within traffic. Moving forward, our focus remains on refining our methods to deliver invaluable insights that traditional sensors and techniques cannot provide.Finally, we aim to provide added value to our clients through an online service delivered as software as a service (SaaS). This platform will empower our clients to visualize the collected data, access results, and glean insights seamlessly. It is a proactive step towards enhancing user experience and facilitating informed decision-making based on data analytics.Specific information to the ChallengeOur goal is to explore the potential of developing a user-friendly analytics platform for clients to interpret data efficiently and generate insightful reports and visualizations for informed decision-making. The work packages include:WP1 M1:M2 Interviews with clients to better understand their needs and views on the proposed online platformWP2 M2:M5 Development of prototypeWP3a M5:M6 Review and feedback from customersWP3b M5:M6 Adaptation of our pricing strategy with a flexible pricing model and a subscription model.The New Mobility Lab can further help us by acting as an introduction/connection point with new potential clients (cities) and companies that can help us with the development of the online platform.

Project Idea Metadata

Project Idea Name: Drones for Mobility solutions as all-in-one service
Date: 4/2/2024 3:11:06 PM
Administrators:
- Manos Barmpounakis

Project Idea Description

Problem

City mobility is a complex dynamical system with different transport modes that interact and compete for limited space. This dynamic environment directly impacts economic sustainability, environmental concerns, and contributes to delays caused by inadequate infrastructure, high levels of motorization, and ineffective traffic management. The increasing complexity of mobility needs and services necessitates the traffic monitoring and the analysis of multi-sensor data, which traditionally consumes considerable budget and time resources and presents challenges related to data formats, quality assurance, privacy, central integration, and compatibility.

Modern cities are equipped with car-centric equipment that cannot adapt to address their evolving mobility needs. Our solution addresses these pressing challenges by harnessing the power of drone swarms to provide high-quality data and fact-based solutions. By leveraging cutting-edge technology, we offer comprehensive insights and actionable recommendations, enabling businesses and municipalities to navigate the complexities of urban mobility with confidence and efficiency.

Why Drones?

Our vision is to reshape urban mobility by identifying sources of traffic congestion and mobility issues and provide actionable solutions. However, current methods rely on outdated and inefficient equipment, leading to unreliable data and a lack of comprehensive insights into mobility patterns across different locations, times, and transportation modes. This results in insufficient data quality to effectively address congestion issues. Common practices such as manual counting, loop detectors, and CCTV systems are prevalent due to their simplicity, but they are labor-intensive and rely heavily on manual effort.

Moreover, these methods fall short in understanding the broader traffic context and identifying the root causes of congestion. Thus, they fail to provide a complete understanding of traffic dynamics and hinder the development of effective solutions.

Additionally, most of the above are not designed for "soft modes" (bicycles, scooters etc.) or pedestrians even though they are crucial for sustainability. Monitoring mobility from the skies is an ideal solution due to latest advances in interdisciplinary topics; the huge advances in drones’ materials and technologies and secondarily advances in Computer Vision, AI and computational power. This is a strong alternative to traditional monitoring techniques as different transport modes (even soft modes) can be identified.

Our solution

We mix a blend of interdisciplinary expertise ranging from traffic engineering to route optimisation, machine learning and automatic control, through computational tools that enable deployment worldwide.

Our aim is to revolutionise how emerging technologies like drones can accurately monitor the various mobility interactions and provide fact-based solutions to reduce their effect on traffic and the environment. While the use of drones for mobility purposes has previously been discussed, challenges associated with the technology had kept its adaptation in low levels, with the examined use cases focusing on small parts of a road arterial or 1-2 intersections.

Since our incubation period at EPFL, our team has been actively exploring opportunities to integrate drones into existing smart city infrastructure. Our unique experiments with drone swarms have paved the way for valuable contributions to urban planning and traffic management.

Using our optimisation algorithms and by taking into account the drone flight regulations, our team designs the drone flight plans which are executed by professional certified pilots with commercial drones for enhanced safety and standardisation. With their birds-eye view the drones are collecting videos in a privacy friendly way, as no vehicle plates nor faces can be identified. The collected videos are then being analysed using an already trained computer vision algorithm which is then fined tuned based on each city’s needs.

We transform the drone into a multi-sensor for sustainable cities

Our vision is to provide fact-based solutions to cities that aim to become more sustainable. Thus, access to emission data is crucial. Using state-of-the-art models and the high-quality data that are produced through our method we can transform the drone into a multi-sensor that can provide strategic CO2 emissions or noise emissions maps in low-cost, non intrusive way. The integration of drones with cutting-edge estimation models removes the necessity for additional sensors either on the drone itself or within the road infrastructure. Hence, our solution can not only analyze and detect traffic congestion and mobility issues but also provide CO2 and noise emissions data for strategic decision makring for urban planners and policy makers.

Our team

Our team is composed of 5 enthusiastic people with experience in cross-disciplinary topics ranging from traffic engineering and computer vision to optimization, control and data science. All of them are closely associated with EPFL. The members of our team are the following:

Dr. Manos Barmpounakis, CEO and co-founder (EPFL Postdoc)
Prof. Nikolas Geroliminis, Scientific Advisor and co-founder (EPFL Full Professor)
Dr. Taylor Mordan, Senior Computer Vision Engineer (EPFL PhD and EPFL Postdoc)
Dr. Dimitris Tsitsokas, Senior Data Scientist and Transportation Engineer(EPFL PhD and EPFL Postdoc)
Mr. Jasso Espadaler Clapés, Senior Data Scientist and Transportation Engineer (EPFL MSc – Best average grade 2021 class)

Professor and co-founder Nikolas Geroliminis is a world renowned researcher with more than 15 years of experience in urban traffic networks and currently serves as Editor-In-Chief of Transportation Research part C: Emerging Technologies journal. Thanks to our EPFL background, the team accounts for numerous scientific publications in top-quality scientific journals. In addition, our advisory board consists of Simon Johnson (mentor and investor in Swiss tech start-ups and advisor to the Swiss National Centre for Robotics) and Dr. Frédéric Dreyer (mentor and coach for start-ups and SME, accredited by Innosuisse and Cantonal Platinn institutions).

Finally we should note that our EPFL spin-off is part of the “Initial Coaching supported by Innosuisse”.

Future Steps

Our team is dedicated to advancing the technology we have developed, currently at TRL levels 7-8, by integrating data from additional sensors such as loop detectors and cameras. By continuously assessing the accuracy of the data collected and estimated, we envision to fast-track the integration of drones into Intelligent Transportation Systems (ITS) infrastructure. Our goal is to offer a cost-effective, efficient, and privacy-conscious multi-sensor solution. Leveraging this technology, we're already making strides in identifying interactions between "soft modes" and pedestrians within traffic. Moving forward, our focus remains on refining our methods to deliver invaluable insights that traditional sensors and techniques cannot provide.

Finally, we aim to provide added value to our clients through an online service delivered as software as a service (SaaS). This platform will empower our clients to visualize the collected data, access results, and glean insights seamlessly. It is a proactive step towards enhancing user experience and facilitating informed decision-making based on data analytics.

Specific information to the Challenge

Our goal is to explore the potential of developing a user-friendly analytics platform for clients to interpret data efficiently and generate insightful reports and visualizations for informed decision-making.

The work packages include:

WP1 M1:M2 Interviews with clients to better understand their needs and views on the proposed online platform

WP2 M2:M5 Development of prototype

WP3a M5:M6 Review and feedback from customers

WP3b M5:M6 Adaptation of our pricing strategy with a flexible pricing model and a subscription model.

The New Mobility Lab can further help us by acting as an introduction/connection point with new potential clients (cities) and companies that can help us with the development of the online platform.

MobiLysis provides mobility solutions as all-in-one service based on drone data

Through this challenge, we aim to start developing an online service delivered as software as a service (SaaS) to our clients and provide added value to our current business model. This platform will empower our clients to visualize the collected data, access results, and glean insights seamlessly. It is a proactive step towards enhancing user experience and facilitating informed decision-making based on real-time analytics.