David Silva
David Silva
David Silva
David Silva
David Silva
David, you work in Wingcopter’s AI and robotics team. What exactly is your role?
I am an AI Software Engineer working in the AI team. I work primarily on providing our drones with the capability to detect other objects using cameras and neural networks. On a daily basis, I develop the object detection model, evaluation metrics, run experiments, and deploy the trained models on an AI edge device.
I also keep an eye on the academic research surrounding this topic, looking for new ideas that I can experiment with to maximize performance and/or minimize inference time.
David, you work in Wingcopter’s AI and robotics team. What exactly is your role?
I am an AI Software Engineer working in the AI team. I work primarily on providing our drones with the capability to detect other objects using cameras and neural networks. On a daily basis, I develop the object detection model, evaluation metrics, run experiments, and deploy the trained models on an AI edge device.
I also keep an eye on the academic research surrounding this topic, looking for new ideas that I can experiment with to maximize performance and/or minimize inference time.
David, you work in Wingcopter’s AI and robotics team. What exactly is your role?
I am an AI Software Engineer working in the AI team. I work primarily on providing our drones with the capability to detect other objects using cameras and neural networks. On a daily basis, I develop the object detection model, evaluation metrics, run experiments, and deploy the trained models on an AI edge device.
I also keep an eye on the academic research surrounding this topic, looking for new ideas that I can experiment with to maximize performance and/or minimize inference time.
David, you work in Wingcopter’s AI and robotics team. What exactly is your role?
I am an AI Software Engineer working in the AI team. I work primarily on providing our drones with the capability to detect other objects using cameras and neural networks. On a daily basis, I develop the object detection model, evaluation metrics, run experiments, and deploy the trained models on an AI edge device.
I also keep an eye on the academic research surrounding this topic, looking for new ideas that I can experiment with to maximize performance and/or minimize inference time.
David, you work in Wingcopter’s AI and robotics team. What exactly is your role?
I am an AI Software Engineer working in the AI team. I work primarily on providing our drones with the capability to detect other objects using cameras and neural networks. On a daily basis, I develop the object detection model, evaluation metrics, run experiments, and deploy the trained models on an AI edge device.
I also keep an eye on the academic research surrounding this topic, looking for new ideas that I can experiment with to maximize performance and/or minimize inference time.
Can you elaborate a bit on what you and the team are working on? Why is this important for the future of Wingcopter and maybe even unmanned aviation in general?
We develop a Detect And Avoid (DAA) system for the Wingcopter 198. The system is broken down into several components: object detection, tracking, collision prediction, trajectory planning, and actuation.
Several complex tasks enable these components to be developed: data collection, labeling, hardware selection, integration of the DAA system in the drone, etc.
To build the system we need people with several different backgrounds and expertise. We need experts in AI, data science, classical algorithms, and every traditional software engineering role.
The DAA system will allow our drones to operate Beyond Visual Line Of Sight (BVLOS), which translates into longer and more autonomous flights. Wingcopter is working at the edge of what is commercially possible to deliver this drone and system, there is no template for it, and regulation is being developed in parallel.
Can you elaborate a bit on what you and the team are working on? Why is this important for the future of Wingcopter and maybe even unmanned aviation in general?
We develop a Detect And Avoid (DAA) system for the Wingcopter 198. The system is broken down into several components: object detection, tracking, collision prediction, trajectory planning, and actuation.
Several complex tasks enable these components to be developed: data collection, labeling, hardware selection, integration of the DAA system in the drone, etc.
To build the system we need people with several different backgrounds and expertise. We need experts in AI, data science, classical algorithms, and every traditional software engineering role.
The DAA system will allow our drones to operate Beyond Visual Line Of Sight (BVLOS), which translates into longer and more autonomous flights. Wingcopter is working at the edge of what is commercially possible to deliver this drone and system, there is no template for it, and regulation is being developed in parallel.
Can you elaborate a bit on what you and the team are working on? Why is this important for the future of Wingcopter and maybe even unmanned aviation in general?
We develop a Detect And Avoid (DAA) system for the Wingcopter 198. The system is broken down into several components: object detection, tracking, collision prediction, trajectory planning, and actuation.
Several complex tasks enable these components to be developed: data collection, labeling, hardware selection, integration of the DAA system in the drone, etc.
To build the system we need people with several different backgrounds and expertise. We need experts in AI, data science, classical algorithms, and every traditional software engineering role.
The DAA system will allow our drones to operate Beyond Visual Line Of Sight (BVLOS), which translates into longer and more autonomous flights. Wingcopter is working at the edge of what is commercially possible to deliver this drone and system, there is no template for it, and regulation is being developed in parallel.
Can you elaborate a bit on what you and the team are working on? Why is this important for the future of Wingcopter and maybe even unmanned aviation in general?
We develop a Detect And Avoid (DAA) system for the Wingcopter 198. The system is broken down into several components: object detection, tracking, collision prediction, trajectory planning, and actuation.
Several complex tasks enable these components to be developed: data collection, labeling, hardware selection, integration of the DAA system in the drone, etc.
To build the system we need people with several different backgrounds and expertise. We need experts in AI, data science, classical algorithms, and every traditional software engineering role.
The DAA system will allow our drones to operate Beyond Visual Line Of Sight (BVLOS), which translates into longer and more autonomous flights. Wingcopter is working at the edge of what is commercially possible to deliver this drone and system, there is no template for it, and regulation is being developed in parallel.
Can you elaborate a bit on what you and the team are working on? Why is this important for the future of Wingcopter and maybe even unmanned aviation in general?
We develop a Detect And Avoid (DAA) system for the Wingcopter 198. The system is broken down into several components: object detection, tracking, collision prediction, trajectory planning, and actuation.
Several complex tasks enable these components to be developed: data collection, labeling, hardware selection, integration of the DAA system in the drone, etc.
To build the system we need people with several different backgrounds and expertise. We need experts in AI, data science, classical algorithms, and every traditional software engineering role.
The DAA system will allow our drones to operate Beyond Visual Line Of Sight (BVLOS), which translates into longer and more autonomous flights. Wingcopter is working at the edge of what is commercially possible to deliver this drone and system, there is no template for it, and regulation is being developed in parallel.
AI is a very broad field. What are the most interesting and what are the most challenging tasks when developing an AI system here at Wingcopter?
I started my journey in AI because in my academic years I specialized in computer vision and unmanned robotics. Neural networks changed the computer vision state-of-the-art completely and by extension, they changed unmanned robotics too.
At Wingcopter, we want to maximize the quality of the predictions of our object detection model while being subject to tight requirements in inference speed, restrictions in computational resources, and complying with strict safety regulations. As stated previously, there is no template we can follow to achieve this. We can be (one of) the first to achieve it - and we are on e very good path. This is both interesting and challenging.
Another challenge, that anyone working in AI faces, is building the dataset...it is hard and expensive :)
AI is a very broad field. What are the most interesting and what are the most challenging tasks when developing an AI system here at Wingcopter?
I started my journey in AI because in my academic years I specialized in computer vision and unmanned robotics. Neural networks changed the computer vision state-of-the-art completely and by extension, they changed unmanned robotics too.
At Wingcopter, we want to maximize the quality of the predictions of our object detection model while being subject to tight requirements in inference speed, restrictions in computational resources, and complying with strict safety regulations. As stated previously, there is no template we can follow to achieve this. We can be (one of) the first to achieve it - and we are on e very good path. This is both interesting and challenging.
Another challenge, that anyone working in AI faces, is building the dataset...it is hard and expensive :)
AI is a very broad field. What are the most interesting and what are the most challenging tasks when developing an AI system here at Wingcopter?
I started my journey in AI because in my academic years I specialized in computer vision and unmanned robotics. Neural networks changed the computer vision state-of-the-art completely and by extension, they changed unmanned robotics too.
At Wingcopter, we want to maximize the quality of the predictions of our object detection model while being subject to tight requirements in inference speed, restrictions in computational resources, and complying with strict safety regulations. As stated previously, there is no template we can follow to achieve this. We can be (one of) the first to achieve it - and we are on e very good path. This is both interesting and challenging.
Another challenge, that anyone working in AI faces, is building the dataset...it is hard and expensive :)
AI is a very broad field. What are the most interesting and what are the most challenging tasks when developing an AI system here at Wingcopter?
I started my journey in AI because in my academic years I specialized in computer vision and unmanned robotics. Neural networks changed the computer vision state-of-the-art completely and by extension, they changed unmanned robotics too.
At Wingcopter, we want to maximize the quality of the predictions of our object detection model while being subject to tight requirements in inference speed, restrictions in computational resources, and complying with strict safety regulations. As stated previously, there is no template we can follow to achieve this. We can be (one of) the first to achieve it - and we are on e very good path. This is both interesting and challenging.
Another challenge, that anyone working in AI faces, is building the dataset...it is hard and expensive :)
AI is a very broad field. What are the most interesting and what are the most challenging tasks when developing an AI system here at Wingcopter?
I started my journey in AI because in my academic years I specialized in computer vision and unmanned robotics. Neural networks changed the computer vision state-of-the-art completely and by extension, they changed unmanned robotics too.
At Wingcopter, we want to maximize the quality of the predictions of our object detection model while being subject to tight requirements in inference speed, restrictions in computational resources, and complying with strict safety regulations. As stated previously, there is no template we can follow to achieve this. We can be (one of) the first to achieve it - and we are on e very good path. This is both interesting and challenging.
Another challenge, that anyone working in AI faces, is building the dataset...it is hard and expensive :)
What do you like about working for Wingcopter?
On the technical side, I like the challenge - delivering drones with a DAA system capable of BVLOS flight. There is a lot of interesting work to do, many skills and subjects to learn, and many innovations to create.
On the social side, there are several aspects that I like:
- Young and international
- Everyone is friendly and supportive
- Everyone is doing their best to design and produce the best drones
- Planned and unplanned events where we socialize and eat together
It makes me happy that some of our drones are used to carry medical supplies that make a real difference in someone's life.
What do you like about working for Wingcopter?
On the technical side, I like the challenge - delivering drones with a DAA system capable of BVLOS flight. There is a lot of interesting work to do, many skills and subjects to learn, and many innovations to create.
On the social side, there are several aspects that I like:
- Young and international
- Everyone is friendly and supportive
- Everyone is doing their best to design and produce the best drones
- Planned and unplanned events where we socialize and eat together
It makes me happy that some of our drones are used to carry medical supplies that make a real difference in someone's life.
What do you like about working for Wingcopter?
On the technical side, I like the challenge - delivering drones with a DAA system capable of BVLOS flight. There is a lot of interesting work to do, many skills and subjects to learn, and many innovations to create.
On the social side, there are several aspects that I like:
- Young and international
- Everyone is friendly and supportive
- Everyone is doing their best to design and produce the best drones
- Planned and unplanned events where we socialize and eat together
It makes me happy that some of our drones are used to carry medical supplies that make a real difference in someone's life.
What do you like about working for Wingcopter?
On the technical side, I like the challenge - delivering drones with a DAA system capable of BVLOS flight. There is a lot of interesting work to do, many skills and subjects to learn, and many innovations to create.
On the social side, there are several aspects that I like:
- Young and international
- Everyone is friendly and supportive
- Everyone is doing their best to design and produce the best drones
- Planned and unplanned events where we socialize and eat together
It makes me happy that some of our drones are used to carry medical supplies that make a real difference in someone's life.
What do you like about working for Wingcopter?
On the technical side, I like the challenge - delivering drones with a DAA system capable of BVLOS flight. There is a lot of interesting work to do, many skills and subjects to learn, and many innovations to create.
On the social side, there are several aspects that I like:
- Young and international
- Everyone is friendly and supportive
- Everyone is doing their best to design and produce the best drones
- Planned and unplanned events where we socialize and eat together
It makes me happy that some of our drones are used to carry medical supplies that make a real difference in someone's life.
Tell us something about your team! What is special about it? How do you guys work together?
All the social aspects mentioned above are also present in our team. Given the scope of the DAA system we naturally have different expertise, and I think we complement each other very well.
Everyone is looking to learn new things, and I value that very much.
Tell us something about your team! What is special about it? How do you guys work together?
All the social aspects mentioned above are also present in our team. Given the scope of the DAA system we naturally have different expertise, and I think we complement each other very well.
Everyone is looking to learn new things, and I value that very much.
Tell us something about your team! What is special about it? How do you guys work together?
All the social aspects mentioned above are also present in our team. Given the scope of the DAA system we naturally have different expertise, and I think we complement each other very well.
Everyone is looking to learn new things, and I value that very much.
Tell us something about your team! What is special about it? How do you guys work together?
All the social aspects mentioned above are also present in our team. Given the scope of the DAA system we naturally have different expertise, and I think we complement each other very well.
Everyone is looking to learn new things, and I value that very much.
Tell us something about your team! What is special about it? How do you guys work together?
All the social aspects mentioned above are also present in our team. Given the scope of the DAA system we naturally have different expertise, and I think we complement each other very well.
Everyone is looking to learn new things, and I value that very much.