Researchers at Luleå University of Technology will investigate how AI can improve the understanding and interpretation of information during forest inspections with the help of drones.
Many areas of use in forestry
Inspecting forests using drones is already being done, but analyzing the collected imagery using artificial intelligence is not as developed. Self-flying drones will become increasingly common in the future, especially in inspection assignments and when it can otherwise be difficult to get around, for example in dangerous environments. The fact that the drones are fully automated leads to higher efficiency, reduced costs and minimized risks of personal injury.
Using autonomous drones in the forest industry in particular has many advantages. It can be anything from getting help in categorizing trees to detecting forest fires. The drones can detect tree diseases, such as infestations by bark beetles, which makes it possible to quickly remedy the further spread of pests in healthy forests. Another area of use is to be able to warn of winds and storm-felled trees that, for example, block roads.
The work, which takes place within the framework of an EU project, is a collaboration between the Robotics Group at Luleå University of Technology and the Skellefteå-based company Conifer Vision, a company that works with AI strategies for image analysis in forestry, renewable energy and the mining industry.
The robotics group, which will primarily carry out the work, is world-class in terms of robotics and drone technology. They work to demonstrate the use of robots under real conditions and fully connected to real needs. The research group has extensive experience of field robotics, ie robots operating in realistic environments outside laboratories and in their work they will use the latest technology in computer vision and location of drones.
Optimize forest management
The aim of the work is, among other things, to evaluate aerial photos and look at how to use AI in forest inspection.Roger Öhlund, CEO of Conifer Vision, says:
– Today, ever higher demands are placed on forestry from a climate perspective. The system we are now developing will be autonomous and self-learning and contribute to objective and quantitative assessments to identify and optimize necessary forest measures for forest asset management.
Artificial intelligence can help create discrimination. That is why the Machine Learning Group at Luleå University of Technology works actively to ensure that future AI solutions combat inequalities and improve gender equality.
In order for artificial intelligence, AI, to be able to act independently, it is trained in making different choices with the help of previous data. In order for AI not to maintain or even reinforce gender stereotypes, the system must be programmed correctly. On the one hand, there must be diversity and equality among those who code, and on the other hand, it is important to have a great awareness of the issues. All of this is something that the machine learning group is constantly working on.
Equal AI systems
Foteini Simistira Liwicki, senior lecturer in machine learning, says:
- In our group, we tried from the beginning to have as many women as men and to treat everyone in an equal way.
György Kovács, postdoctoral fellow in the same department, agrees:
- We work hard to ensure that men and women have the same opportunities and are treated equally. The benefits of gender equality are many, it provides a better work environment, more job satisfaction and better results.
Develops AI technology for humans
Gender equality work is also reflected in the work towards the companies and in the projects they work in, including the AI project Applied AI DIH North, which includes a number of company cases. AI is dependent on data, lots of data. Algorithms model real data, and if gender inequality exists in society, it also exists in the data. Thus, there is a great risk that AI will also be unequal. As a result, developers may unknowingly add gender prejudices to AI systems.
Richa Upadhyay, PhD student in the group says:
- This is really not the kind of development AI predicts. Therefore, it is just as important to develop an understanding of gender equality in our societies as it is to develop an understanding of our AI systems.
Foteini Simistira Liwicki, continues:
-Because we develop AI technology for humans, it is important that the developed technology has the different opinions of all people and that it is not biased with regard to gender. Gender equality requires that both women and men are equally involved.
The field of computer science has been male-dominated for decades. It's changing, but it's slow. However, Foteini is hopeful.
-I am the educational leader for the new 5-year engineering program Applied AI and for the first time, a large number of women have applied for the program. It makes me so happy for our future in AI!
Researchers at Luleå University of Technology will, together with the company Substorm, investigate whether it is possible with the help of AI to increase trust in digital communication channels.
Predict, analyze and improve trust
In today's society, communication and conversations have largely been moved to digital channels. These are often short messages such as emails, chats or tweets.
Before we trust a person or an organization, we make both conscious and unconscious assessments based on personal criteria and experiences. But much of this assessment is done through non-verbal communication during physical meetings. What then does it look like when trust is to a greater extent to be built via shorter text messages via digital meetings instead of through human meetings?
By collaborating with Luleå University of Technology in an AI project, the company Substorm hopes to discover ways in which AI can be used to predict, analyze and improve trust in digital communication channels based on text. The company has been ranked as one of Sweden's most innovative AI companies. They were founded in Luleå in 2019 and their business spans areas such as intelligent automation, robotics and machine learning.
Model must detect changes and intervene
Niklas Karvonen, CTO at Substorm, participates in the project. He says:
- Trust is an important factor in our everyday lives. Before a person decides that they can trust someone or an organization, an assessment is made based on a number of criteria. Communication is a central part of building trust, and a large part of human communication is non-verbal.
The goal of the work is now to develop a model that will be able to predict what trust a person or an organization will receive and see if it is possible to predict a possible change in trust over time. This would also make it possible to intervene when confidence in a communications partner declines.
AI is crucial to achieving a sustainable and equal world
Norrbotten is growing strongly and the need for competence in machine learning is increasing in the region. Both Luleå University of Technology and Substorm have built strong teams in the area, and also managed to balance the distribution of men and women in a good way.
Substorm, founded by a woman, is convinced that AI will be crucial to reaching a sustainable world. Hiring 50/50 women and men is not something they compromise on or strive for, it is simply a must. This means that the company works hard to find expertise in forums, organizations and networks where they know that there is a high degree of women and non-binary people, such as Women in AI Sweden and Women in RPA. They also work with role modeling, which means that women have key management and technical positions to attract more people.
Technology knowledge must improve society
Niklas Karvonen believes that it is important for them to work ethically and sustainably. He says:
- We are happy to get involved in projects where we see that our technical knowledge in e.g. deep learning can be used for a better society, e.g. the research project NoBias for non-exclusive communication, or AboutMom which is building an app for better postpartum care.
The work in the AI project Applied AI DIH North is in line with these goals. They also give them opportunities to further strengthen ties with the university, where they already collaborate through degree projects, and as part of project courses.
AI researchers will make it easier to create photo books
Welcome to a three-day worldwide online event around artificial intelligence!
Learn more about the Swedish strategy, the national AI ecosystem and its challenges. Listen to thought leaders from the industry, public sector and academia talk about key AI focus areas.
Read more: https://swedeninnovationdays.confetti.events/
Making sense of business models – a webinar for practitioners
Are you interested in learning more about business models? Curious to know how digitalization is leveraged for business impact? Or perhaps you want to understand the challenges and opportunities of business model innovation? If you answered yes to any of the above, grab your spot at our upcoming lunch-time webinar!
In our 45 minute webinar taking place on 5 November 2020 @ 12:00 (CET), we guide you through the basics of business models in B2B contexts and highlight the opportunities presented by digitalization, aiming to provoke discussion and provide new insights. This initiative is driven by a desire to demystify business models and promote a common understanding to support industrial practitioners in their day-to-day work.
Register here: https://ltu-se.zoom.us/meeting/register/u5ApfuyhqjwsH9RHraQl6MQXr6m7jsNV2scd
Kurs: Building AI II
Building AI is for anyone who wants to improve their AI-related vocabulary and skills, including non-programmers and people who can program in Python.
By taking the course, you'll learn more about what makes different AI methods possible and where and how these methods can be applied in real life. As a result of this course, you will be able to craft your own AI idea and present it in an understandable format.
Läs mer här: https://buildingai.elementsofai.com
Process IT Innovations is together with the sister organization Automation Region and the strategic innovation program Processindustriell IT och Automation (PiiA) the initiator of the work to become a national candidate for EDIH – European digital innovation hubs.
The Robotics and AI research group continues its successful journey in NASA/JPL’s team CoSTAR in the Subterranean Challenge, the biggest robotics competition in the world. The team won the second circuit of the competition and is now preparing for the third and last overall combined challenges.
When: 29 - 30 Oct. 2020, 09:00
The 8th Swedish Workshop on Data Science (SweDS20) will be hosted by the Luleå University of Technology (LTU), in 2020 on October 29 and 30, in Luleå, Sweden.
SweDS is a national event with a focus of maintaining and developing Swedish data science research and its applications by fostering the exchange of ideas and promoting collaboration within and across disciplines. This yearly workshop brings together researchers and practitioners of data science working in a variety of academic, commercial, industrial, or other sectors. Past workshops has included presentations from a variety of domains, e.g., computer science, linguistics, economics, archaeology, environmental science, education, journalism, medicine, health-care, biology, sociology, psychology, history, physics, chemistry, geography, forestry, design, and music. SweDS20 is organized by the Luleå University of Technology (LTU).
On October 1, a webinar, a digital tour in the 5G test environment, was conducted at Luleå University of Technology. During the event, the participants were shown around the environment and received information about exciting 5G applications. A number of pre-recorded films were shown with interviews with researchers, companies and students.
Project manager Arne Gylling talked, among other things, about how it is done when you as a company come and test in the environment. A 360-degree camera showed around the environment both outside and inside.
One of the projects that participated in arranging the webinar was the AI DIH North project. Artificial intelligence and 5G are technologies that reinforce each other. 5G transports data and information, enables huge amounts of data to be sent at extremely high speeds to unimaginably many devices, while artificial intelligence analyzes and offers opportunities for more efficient solutions. An example of this is the robotics group at the university that flies autonomous drones connected to 5G. This means that as soon as the drone has left the ground, there is no pilot in control. The drone is programmed to use AI to analyze its surroundings and make its own decisions, for example to be able to independently avoid obstacles. Autonomous drones can be used in many areas, for example to carry out risky inspections such as in mines. The robotics group is world-class in terms of robotics and drone technology, has won prestigious awards and collaborates with NASA.
- With the help of robotics, intelligent automation and simulation, very difficult and dangerous tasks can be performed. This is a growing subject area, both nationally and internationally. Robotics and artificial intelligence, for example, have been identified by the European Commission as one of the most important subject areas in the Horizon Europe 2021–2017 research initiative, said George Nikolakopoulos, professor of robotics and automation.
One of the companies that is involved and will conduct tests in the 5G environment is Compodium. It is a Luleå based company that delivers services in digital video conferencing and was the first in the world to provide secure digital care rooms. Bengt Grahn, founder of the company, said that he thinks it is super exciting to be involved and put his services on the latest radio technology.
Finally, the student Jesper Gladh was interviewed, who is now doing his ex-job in 5G and Virtual Reality at Ericsson.
The 5G environment is open and welcomes organizations to test products and services and is run together with Ericsson, Telia and Tieto Evry. The webinar was part of the EU's digital information campaign "My Europe", which is about disseminating information about EU-funded projects.
The event was organized by the projects:
Wireless Innovation Arena
Applied AI DIH North
5G for health and care in Upper Norrland
Arctic 5G Test Network
Wednesday, May 20, 2020, 08:00 – 08:50, Zoom
Welcome to the first AIM Point, a series of focused webinars about different hot AI topics of today. The AIM Point is an activity within the Applied AI DIH North project. First topic is about Chatbots!
8.00 - 8.05 Michael Nilsson, Applied AI DIH North project manager, will introduce the seminar.
8.05 - 8.10 Examples of challenges regarding AI at a municipality by Ola Lidström, Process leader digitization, Piteå Municipality
8.10 - 8.40 Marcus Liwicki, Professor of Machine Learning at Luleå University of Technology, will talk about what a chatbot is, state of the art, what do we miss today, research challenges and future work.
8.40 - 8.50 A Q&A session will follow.
For those of you who want to discuss more in detail stay, and/or let us know so we can contact you afterwards.
About Marcus Liwicki
Marcus Liwicki was born in 1982 in Berlin. In 2007, he defended his dissertation at the University of Bern in Switzerland, with a dissertation on automated handwriting recognition on whiteboard. He has been awarded both the CEBIT Innovation Award 2015 and the Young Investigator Award 2015. Marcus Liwicki presented his doctoral dissertation six years after he finished high school. At the age of 29, he then became the youngest professor of computer science in Germany. Today his research is largely about pattern recognition, on both a large and a small scale. It may be about loading satellite images to find geographical locations on other planets or teaching computers to read writing and analyze the text. Currently he is chaired professor at Luleå University of Technology and a senior assistant in the University of Fribourg. Marcus Liwicki is a co-author of the book "Recognition of Whiteboard Notes – Online, Offline, and Combination", published by World Scientific in October 2008 and he has more than 200 publications.
Read more about Chatbots here:
Register to the event here:
Link to the Zoom webinar here:
Strengthened innovations, greater competitiveness and more sustainable development in applied research in artificial intelligence. That is the purpose of a new project, Applied AI DIH North, which is led by Luleå University of Technology.
Everything indicates that artificial intelligence's importance in society will increase in the future. For example, AI will change our way of working, new jobs will be created and less attractive jobs will be gone. In the long run, the entire society will be permeated with products and services linked to AI, a fact not least relevant to the northern region where Luleå University of Technology is located, a region that is partly defined by sparsely populated areas and an aging population.
Important functions in society and business' competitiveness are dependent on how innovation can create benefits for Sweden and the region. The project Applied AI DIH North will contribute to competitive companies and make the region more attractive for start-ups. The result will be long-term sustainable growth that attracts people, investments and companies to the region.
– Norrbotten must take a leading position in the AI area to support continued positive development, says Michael Nilsson, project manager at the Centre for distance-spanning technology at Luleå University of Technology.
– For competitive reasons, our region's small and medium-sized companies need support in the AI area. It is important to maintain a high level of expertise with leading industries that are important and relevant to the outside world.
Luleå as AI-hub
For Luleå University of Technology, the project is a logical complement to the AI research that is already underway at the university at the AI Excellence Centre. In particular, the university has expertise in applied AI. Through collaboration with companies, research is translated into innovations for the benefit of industry, education and health.
– A project of this kind suits us very well. We have good technical infrastructure, several different lab environments and existing test beds. In addition, our research is one of Sweden's most applied, we have a very strong collaboration with industry and a high degree of external research funding. All in all, we are extremely suitable to be a hub in AI-related research, says Michael Nilsson.
– However, our overall goal is to reach beyond regional and national levels. If we are to be really attractive, only world-class counts.
Copyright © 2021