Scientific program

Dec 07-08, 2022    Chicago, USA

6th International Conference on

Robotics, Machine Learning and Artificial Intelligence

  • Home -
  • Scientific program

Keynote Form

Ioana Triandaf

Ioana Triandaf

Navel Research Laboratory, USA USA

Title: Delay induced swarm pattern bifurcations in mixed reality experiments

Abstract:

Statement of the Problem:  Natural swarms exhibit patterns in a variety of forms and have inspired researchers to understand how simple organisms produce complex, emergent patterns occurring when individual organisms follow simple dynamics and local rules. Our work provides a model for the swarming behavior of coupled mobile agents with a communication-time delay which exhibits multiple dynamic patterns in space, which depend on interaction strength and communication delay. Methodology & Theoretical Orientation: A thorough bifurcation analysis has been carried out to explore parameter regions where various patterns occur. We extend this work to robotics applications by introducing a mixed-reality framework in which real and simulated robots communicate in real-time creating the self-organized states predicted by the theory. The mixed-reality framework allows for the systematic and incremental introduction of real-world complexity by coupling a few real robots and a large number of idealized (virtual) robots together in a swarm - the latter being well understood. Findings: The proposed swarm controller was tested on two different robotic platforms: NRL’s autonomous air vehicles and UPENN’s micro-autonomous surface vehicles on water. Theoretical pattern formation results are confirmed in mixed-reality experiments. Conclusion & Significance: Increased understanding of challenges for real robots is obtained as a systematic, incremental verification of swarming behavior at low cost and risk of damage.  Switching between patterns is achieved in the hybrid experiments, thus simulating the flexible behavior of the real robotic system.

Biography:

Ioana Triandaf is an applied mathematician specializing in dynamical systems and numerical methods for partial differential equations. She has been modeling swarms since 2004. She is the recipient of the NRL 2005 Alan Berman award for her work on swarming. Since 2017 she collaborated with roboticists in implementing swarming motion on robotic systems. Currently, Triandaf is focusing on analyzing and testing swarm disruption methods and metrics.

Tomasz Bak

Tomasz Bak

Digital Fingerprints, Katowice, Poland Poland

Title: From Internet access devices usage to behavioural model

Abstract:

The way of using a mouse, keyboard, tablet, or phone can be a great source of the information about user behavior. Data can be gathered from a spectrum of sensors built into these devices. Then this data can be transferred into features that characterize users, like manner of typing, speed of cursor, or mouse movement curve. This presentation will focus on how to use those features to build behavioral models which will be unique and adjusted to each of thousands of customers. This problem has some important additional aspects. For one user several different models can be built, therefore some aggregation methods have to be delivered. Some of the data have to be anonymized. Continuous authentication is excepted by businesses. These factors define important limitations for model productionization. They will also be discussed.

Biography:

Tomasz Bąk has completed his Ph.D. in 2018 from the University of Economics in Katowice in the area of spatial sampling. He is the Head of AI in Digital Fingerprints, a company that provides a tool that would defend every user of the critical service from the digital hijack. He has published several papers in reputed journals and in the same gathered 8 years of experience in developing ML & Data Science solutions for business.

Amel Mhamdi

Amel Mhamdi

FSEG University of Tunisia, Tunisia France

Title: Projection of the corpus for the speech recognition service from standard French to French regional accents in field of French media

Abstract:

Many companies, and in particular IT services and engineering companies, are faced with the challenge of automated research and above all reliable application of Curriculum Vitae (CV), in multiple formats. JEMSDoc is a prototype of a search engine based on a big number of resumes. It much this CV with a specific mission. Our prototype respects increasingly stringent security constraints in terms of pseudonymization, encryption, and consent of individuals within the framework of the regulations in force: The General Data Protection Regulation (GDPR). to meet the needs of our human resources team, we used a service-oriented architecture that starts with the first API for data analysis followed by another for the implementation of a classification model. But beyond the technical knowledge of data analysis and artificial intelligence, we will proceed with a Design thinking approach through an evolutionary process that stimulates innovation. With this strategy, our solution reflects the user experience well with ergonomic and human-machine interface response time requirements by hiring managers, business engineers, UX design, a product owner, and a data scientists' team.

Biography:

Amel Mhamdi, completed his Ph.D. at the age of 31 years from FSEG University of Tunisia and member of a LIMTIC research laboratory. She has worked from 2016 until 2018 in the banking sector for the establishment of artificial intelligence service in finance innovation projects especially on the subject of Deals and banking reconciliation with Natixis-Paris and on the subject of credit automation with BNP-Paribas. Since 2019, she has been focusing on the mapping and industrialization of IA research and media innovation projects with France-Télévision. She has published more than 7 papers in reputed conferences and journals.

Steven Lockey

Steven Lockey

University of Queensland Business School, Australia Australia

Title: Trust in Artificial Intelligence: What do we know and why is it important?

Abstract:

The rise of Artificial Intelligence (AI) in our society is becoming ubiquitous and undoubtedly holds much promise. However, AI has also been implicated in high-profile breaches of trust or ethical standards, and concerns have been raised over the use of AI in initiatives and technologies that could be inimical to society. Public trust and perceptions of AI trustworthiness underpin AI systems’ social license to operate, and a myriad of company, industry, governmental and intergovernmental reports have set out principles for ethical and trustworthy AI. To guide the responsible stewardship of AI into our society, a firm foundation of research on trust in AI to enable evidence-based policy and practice is required. However, to inform and guide future research, it is imperative to first take stock and understand what is already known about human trust in AI. As such, we undertake a review of 100 papers examining the relationship between trust and AI. We found fragmented, disjointed, and siloed literature with an empirical emphasis on experimentation and surveys relating to specific AI technologies. While findings suggest some convergence on the importance of explainability as a determinant of trust in AI technologies, there are still gaps between conceptual arguments and what has been examined empirically. We urge future research to take a more holistic approach and investigate how trust in different referents impacts attitudinal and behavioral intentions. Doing so will facilitate a more nuanced understanding of what it means to develop trustworthy AI.

Biography:

Steve Lockey is a Postdoctoral Research Fellow in Organisational Trust at the University of Queensland, Australia. He received his Ph.D. from Durham University in 2017. His research interests primarily relate to the development, repair, and measurement of trust in public and private sector settings. Currently, he is investigating the relationship between trust and Artificial Intelligence and the multilevel nature of organizational trust. Steve’s research has informed policy in the United Kingdom, and his scholarly work is published in Business Ethics Quarterly, Personel Review, and the International Journal of Police Science and Management.

Speakers

Alain De Maertelaere

Alain De Maertelaere

Cronos NV, Belgium Belgium

Title: Data-driven decision making - Guiding organizations towards an “AI proof” status

Abstract:

Needless to say that DATA is the most important driver for data-driven digital transformation. In one of its latest reports, Gartner indicates that the number of AI projects will double in 2020. Where today we have 4 to 5 AI projects per company, we go to 35 projects in 2022. The biggest challenges for the implementation of AI are the lack of specialists and, the concerns, and the lack of a clear and effective framework about data quality and defining the scope. Data-driven decision-making is about decisions that are made based on insights gained from (historical) company data by applying data analytics and AI. We forget about intuition, observation, or "informed guesswork": no more shooting in the dark! We say that data is the oil of the 21st century, analytics and AI are the combustion engine. The insights gained from data can be used by companies:
1. to provide valuable information to optimize their current operational efforts and thus become more customer-focused;
2. to forecast future trends;
3. to make them more adaptable to the constant state of change in the digital world;
4. to help them develop strategies and new activities (cf. blue ocean) to generate more revenue.
During the presentation we will explain how we can make the traditional organization AI proof by initiating methods and techniques for the creation of company awareness about the importance of data, the setup of data maturity tracks, the improvement of data quality, "ideation" setup, etc.), and what architectural resources we need to consolidate and store company data to achieve all these and make it accessible to analytics and AI. By the end of the session, it will be obvious to the listener how decisive and how important it is to have a robust, consistent data foundation layer for obtaining reliable results through analytics and AI.

Biography:

Alain De Maertelaere has been working as a freelance consultant for more than 25 years in the domain of data warehousing, business intelligence, data analytics, and artificial intelligence. Due to his experience and expertise, he has been interviewed by many IT-related magazines in Belgium. In 2016 he has awarded by a Belgian customer for initiating and implementing a “data warehouse dream team”. In 2017, Alain has been an award winner of “Innovation by technology”.

Leandro Romualdo da Silva

Leandro Romualdo da Silva

IBM, Brazil Brazil

Title: Artificial intelligence applied in business operations

Abstract:

Actually, many business areas use artificial intelligence because of the growing power computational and the of amount data generated in the last years is very representative of the current moment of technology. Nowadays already have multiple machine learning models applied in companies, but now with AI and RPA we can do more for the companies, and AI and RPA make the companies be more competitors with automatization internal processes using computer vision and PNL. New tools for PNL and computer vision can contribute to accelerating the development of new solutions, this brings agility for companies' areas. For example accountants, tax, financial, legal, and other areas. I’m going to talk about how some companies become more competitive using AI and RPA in diverse areas.

Biography:

Leandro is a degree in databases and business intelligence for FIAP in São Paulo – Brazil. Have more than ten years of job experience with data and AI, and now is an Artificial Intelligence Specialist in IBM Brazil, have a great experience with data and AI projects around entertainment, telecommunications, consulting, schools, e-commerce, insures, risks, government companies, and others. He Published content on websites and magazines in Brazil, was the speaker in IBM events and others events, talking about data, data science, artificial intelligence, and AI applied for business.

Ouardi Amine

Ouardi Amine

Hassan II University, Morocco Morocco

Title: Optimizing Heuristic Search Algorithms using Neural Networks

Abstract:

On the opposite side of the uninformed search algorithms, performing a systematic search, heuristic search algorithms are based on multiple rules leading them to estimate, in a predictive way, the minimal cost of the path from the current state to the goal. In this sense, the A* algorithm is an example of heuristics-based algorithms that can guarantee to find a least-cost path to a goal state if this algorithm is using an “admissible heuristic”. A heuristic is said to be “admissible” if it never overestimates the real path cost from the current state to the goal. Furthermore, if the condition h(x) ≤ d(x, y) + h(y) is satisfied by the heuristic h (d denotes that edge length), for every edge (x,y), then his called consistent. And with consistent heuristics, finding an optimal path without processing any node more than once is guaranteed. The main idea consists of developing a Neural Network that can optimize those heuristics to further refine the A* algorithm results. Towards achieving that goal we must find the best synaptic coefficients, and for that reason, a learning phase will be needed during which the network parameters are adjusted until the best admissible and consistent heuristic is obtained, dominating any other heuristic (h1 dominates h2 if for every node n (state), h1(n)>h2(n) ). During this learning phase, and as inputs, the neural network will have some representative examples in the form of pairs of several problems and heuristics ({P1,h1};{P2,h2}...{Pn,hn}), to finally be able to calculate the best heuristic regardless of the inputs.

Biography:

Ouardi Amine, 28 years old, head of Architecture perimeter, ELIS project, Capgemini; Artificial Intelligence Ph.D. student at ENSET Mohammedia, working on optimizing heuristics search algorithms using Neural Networks. Had a Master's degree in Imaging and Business Intelligence, with a graduation project on the Internet Of Things, including QR codes and NFC technology. Got a fundamental license degree in Mathematics and Computer Science, with a final project related to Genetic Algorithms: studying optimal solutions for the Travelling Salesman Problem.

Mohamed Sadek

Mohamed Sadek

Com-Iot Technologies Dubai, UAE

Title: The impact of AI on smart cities in the areas of road, transport, traffic & security

Abstract:

Statement of the Problem: A smart city is a safe & secure city for all its inhabitants. A vibrant and dynamic city is full of life and activities on its roads, streets & highways. Also, its residents require security in their homes and workplaces. Applying AI technologies such as machine learning and deep learning on data generated from the city, such as road data, vehicles on the road, road lanes, and with the use of high precision devices such as LiDAR sensor technology and vehicle perception systems can result in powerful capabilities that capture these raw data and transform it to extremely valuable information and data that cities, road companies, enforcement authorities, and municipalities can use to learn details about road conditions, utilization, accidents and events, traffic violations, driver behaviors, and much more. Likewise, in the area of physical security, by applying AI with LiDAR sensors and people perception systems, the system can gather raw data from locations requiring high levels of security such as airports, prisons, international borders, military installations, critical infrastructures, and others leading to applications such as perimeter protection, exit/entrance monitoring, gate access control, restricted zone enforcement, people counting, precise crowd counting, crowd management and much more. Sensor fusion between LiDAR sensors and cameras can only enhance the outcome and provide critical visual evidence of violations, intrusion, and other alert generating incidents which can also be live-streamed as well as video recorded on-demand.

Biography:

Current: CEO/Founder COM-IoT Technologies. A Dubai, UAE-based AI company. Previous: Managing Director of SES_O3B Networks in the MECA region. Upon completing his Master’s of Science degree in Computer Science from the University of Illinois in Chicago in 1989 Mohamed Sadek joined AT&T Bell Labs in Naperville, IL. Where he led a team of software developers working on feature development for AT&T. In 1997 he completed his MBA from the Wharton School of Business, University of Pennsylvania, and led Lucent Technologies in Saudi Arabia until the end of 2000 where he achieved sales of over $1.3 B. In January 2001 He became MD for Nortel Networks in the Middle East until 2005 where he rejoined Lucent Technologies in Abu Dhabi as head of Sales for the Middle East & North Africa. In 2009 Mr. Sadek joined Norconsult in Saudi as VP, International Projects and in April 2013 he joined O3B Networks.