Blue Jay is on the news
Watch Blue Jay on the news of RTL4 on 10-07-2020!
Also, read one of the news articles about our drone:
Blue Jay drone lends a helping hand during fire
A group of students from the Technical University of Eindhoven envisions developing a drone with added social value. The drone functions as an assistant for the fire brigade, which is especially useful and needed in elderly homes.
Last year, 22 people died in a fire in the Netherlands. Over sixty percent of these victims had an age over 65 according to the annual overview of fatal domestic fires in 2019. By detecting the fire in an early stage and assisting the residents and the fire brigade, Blue Jay can save lives!
“An autonomous flying drone is an amazing opportunity to improve fire safety”, explains Paul van Dooren of the Dutch fire brigade. According to Paul, Blue Jay could protect the vulnerable elderly, “A son of an aged man in an elderly home gave his father an electric frying pan. However, the man put the pan on the stove. It caused a fire and the man passed away. Accidents like these could be prevented in the future by quick alarming of Blue Jay.”
Technology for good
On the 9th of July the team presented their proof of concept. In a real-life scenario, the drone assisted a resident of an elderly home during a fire. When a fire breaks out in an elderly home, it is protocol that all the personnel must wait outside until the fire brigade arrives. Due to this, crucial time is lost.
By means of a heat sensor the drone detects the unsafe situation. With this knowledge the drone can determine a safe route for evacuation. Artificial intelligence is used to develop an algorithm detecting people, which is implemented in the drone. Blue Jay can also inspect the building looking for residents in danger. This information is shown to the fire brigade via an app. In this way, the approximate 7 minutes befor the fire brigade arrives will not be lost and be used to help residents out of their homes.
International and multi-disciplined
The Blue Jay team believes that drones and people can cooperate. The on-going project has the ultimate goal of developing an indoor drone that can assist in domestic and healthcare settings. The team consists every year of new ambitious students, setting aside their study to participate in the project. The current Blue Jay team consists of 17 students from the TU/e and Fontys with different majors and 4 different nationalities.
To effectively work together the team is divided in multiple subteams. The Interaction Design subteam makes sure that the interaction between the drone and the user is optimized and that the needs of the user are satisfied. The hardware subteam is looking at aerodynamics, electrical efficiency and mechanics to ensure usability and flyability of the Blue Jay drone. Within the software team the challenge of autonomous flight is tackled and by means of AI multiple detection problems are solved.
Every academic year an entirely different and new Blue Jay team starts. Continuity is assured through an advisory board that includes some of the previous year’s students who support the next year’s team. “In this way it is made sure that the long-term Blue Jay vision is ensured and that every team has its freedom for new insights and innovations.” explains Marieke Vermeulen, PR and finance manager of the team.
Collaboration equals innovation
“Blue Jay would not have been this far without our valuable partners” explains Anne Smits, team manager at Blue Jay Eindhoven. The network of Blue Jay has extended the last couple of years. Not only do the sponsors support the Blue Jay team with equipment and knowledge, but their experience and culture are also of great value to Blue Jay. The TU/e, NXP, Fourtress and Focus are the main project partners of the student team.
At the 60th anniversary of the TU/e, a group of students was asked to develop an innovative technical product. The students came up with the world’s first Drone Café in which a drone prototype served drinks. After this event the students saw the potential of drones in our society. This was the start of the Blue Jay. However, the TU/e is not a main sponsor for this reason only. They support the team financially, with high-quality trainings (e.g. acquisition and project management) and a facility where they can work.
NXP is helping the Blue Jay team by supplying Mobile Robotics technology for Flight Management Units, Companion co-processors and robust communication channels. NXP has a long history of delivering solutions for automotive and aerospace applications. Many of these solutions also apply to the rapidly evolving field of mobile robotics. Whether your system operates on the ground, under the sea or in the sky, NXP offers a complete portfolio of sensors, controllers and communications solutions. Therefore, NXP can provide a large part of the electronics Blue Jay needs as a starting point for their drone developments. For the AI part, Blue Jay was relying on a reference design called ‘Coral board’ combining the NXP i.MX8Q microprocessor and the Google’s Edge TPU.
Learning from each other
NXP and Blue Jay go back for 4 teams already. Former Blue Jay member, Jari van Ewijk, even took the opportunity to start as a working student at NXP in 2018 while continuing his electrical engineering study. Jari is still of great value for the team: “At the start of the academic year I organized a workshop for the Blue Jay team to get them quickly up and running”
Also, the collaboration with Blue Jay is valuable for NXP. The student team is testing some of the new technologies and provide feedback on ease of use and performance. Next to that, a current Blue Jay member just started his NXP internship on companion processors and security.
The current Blue Jay team has been facing extra difficulties due to the COVID-19 crisis. Their working office has been closed and they needed to work remotely. Jari joined them regularly during conference calls focusing on hardware and software to help them through this difficult period. “Even in these times, the team continued and prepared for their final event, which is remarkable.”
Valuable software and management knowledge
Fourtress develops unique and innovative software solutions for technical systems and with a focus on improving lives. From low level embedded software to hardware management, all the way to deploying middleware applications and the development of mobile apps.
“We’ve been involved since the first Blue Jay team and we see it as a win-win. In the beginning we simply provided technical knowledge. But later we also loaned the team a software engineer for a few months and provided a student to work on the Blue Jay mobile app as a thesis project.
This time we helped them improve their focus on the work that needs to be done. We provided support in setting up and configuring tools (Confluence & JIRA) in the Atlassian Cloud. This enabled them with an online team workspace for sharing knowledge and easier collaboration. Furthermore, it allowed them to plan, track and deliver new releases more efficiently.
Since the beginning we see our cooperation as a win-win situation for both. We provide them in technical knowledge, the way of work and experience. But we also learn from them. We like their ambition, energetic way of work and enthusiastic environment.
How we operate with Blue Jay is a bit of a model for how we try to operate with all our customers. In that, if you hire one Fourtress person, you hire all of Fourtress: all our knowledge is at your disposal. Together we go further!”
A crucial part of developing a drone that can assist the fire brigade, is that it needs to fly autonomously. Therefore, Blue Jay and Focus started collaborating. At Focus, a Vention spin-off, the Ultra-Wideband technology is developed. This technology enables both outdoor and indoor positioning. In the Ultra-Wideband system anchors are used that know their own location of which they are able to determine the location of what is tracked, so called ‘tags’. The Blue Jay drone is equipped with two tags to determine its location and direction. Niels Verstappen, software engineer at Blue Jay, “With Focus’ technology we are making indoor flight possible.”
Get in touch with the team if you’re interested in supporting their work, getting involved or becoming a partner. Send an email to: firstname.lastname@example.org. If you want to support the team financially, visit the crowdfunding page: https://crowdfund.tue.nl/project/team-bluejay
A visual representation of Blue Jay’s work can be seen in this video: https://www.youtube.com/watch?v=WuWrWdZTojI&t=39s
How is Jari doing now?
We had a meeting with one of our most involved Blue Jay members of the Blue Jay comminity
What is your name?
I am Jari van Ewijk, and I am 23 years old.
In which Blue Jay team were you?
I was part of Blue Jay 3 between July 2017 and September 2018. My team created the proof of concept for a lifeguard assistant drone. It was able to spot a drowning person in a swimming pool, drop an inflatable buoy and alert a lifeguard.
What was your function within Blue Jay?
I was a full-time team member for a whole year. In the first place I was an electrical engineer, but I was an embedded software engineer as well. And occasionally also a drone pilot.
I was one of the team members responsible for building the drones. We build six in total, though many did not survive until the end of the year. Together with my colleague, I was responsible for designing and manufacturing the circuits of the drones. We also had to select and integrate many subsystems that we could not design ourselves. I also acted as the “negotiator” between the electrical engineering and software development teams.
It was a lot of fun to have these different roles within the team. I was able to learn a lot that I wouldn’t have been able to learn during my courses and projects at the university.
How did you experience your year at Blue Jay?
During my year at Blue Jay I got to work with a group of highly motivated students from different study backgrounds. It was quite an experience to work with them and we shared a lot of fun moments together. It’s always great to reminisce about the amazing things we did together.
I also got to attend a few big events and meet lots of people working for very interesting companies. Even though there is a lot to learn by yourself as a member of a student team, I think I learned just as much from the amazing people that helped us in many ways as part of various sponsorship deals. I learned that there is a lot of knowledge available in the industry, ready to be shared, and it has been a privilege to have access to some of that knowledge.
However, it was not only a fun and educational year. We worked very hard to achieve our goals, but not everything went as planned and we had to learn that the hard way a few times. Especially the setbacks during the last few months were difficult to process, but it is also the reality of life sometimes. In the end we can be proud of what we were able to achieve. Hopefully, our legacy will remain visible to many future team members of Blue Jay.
What are you doing right now? (Still studying, working etc)
I am currently finishing my bachelor’s degree in electrical engineering at Eindhoven University of Technology. I will start with my master’s in September. It took me five years to finish my bachelor’s, two more than originally planned. One year was because of Blue Jay, but then I decided to take another extra year to finish my bachelor’s, because I was offered a part time job at NXP Semiconductors.
I started working at NXP in August 2018, just after I left Blue Jay. At NXP I work about 16 hours per week on projects that are related to mobile robotics. That also includes drones: we are developing products for autonomous and functional safe drones. One of our biggest projects in mobile robotics is the HoverGames. Participants of the HoverGames try to come up with technical solutions to societal challenges using our NXP drone development kit.
NXP is also still a close partner of Blue Jay, so I try to help the team as much as I can. I advise the current team on technical matters and I try to help with troubleshooting when there are problems.
How did your further career benefit from you year at Blue Jay?
In the first place, my year at Blue Jay directly led to me being offered a (part time) job. Being part of a student team is a very valuable experience that will pay itself back, even though we are not getting any (monetary) compensation whatsoever. I got to improve both my technical and soft skills during my year at Blue Jay. A student team is a great environment to develop yourself and learn through making mistakes. Yes, many mistakes were made…
Being part of a student team is also a great opportunity to expand your professional network. You get to know many people that work for all kinds of interesting companies, but also people from the academic world and even fellow students that have joined different student teams. It really surprises me how often I run into people that I was introduced to during my year at Blue Jay.
Blue Jay showed me what it is like to work together in a multidisciplinary team. Blue Jay has always had students from many different backgrounds. I think working together with other people that might have a completely different view of the work is a very valuable thing to learn at the beginning of your career.
How close are you to our drone?
Tijn Dieker & Suzanne Blokland
Blue Jay Eindhoven
Drones have the power to extend our senses and actions. However, the real potential of drones is still subject to experimentation and exploration. Blue Jay Eindhoven is a student team that explores the potential of drones assisting people. Blue Jay focuses on autonomous drones that can cooperate with humans in a safe, interactive, and helpful way. Every component of the drones Blue Jay develops is designed with the user in mind because we want to create a product that can be easily adopted by our users. We do not expect our users to be experienced with drones, so the drone should be easy to use by everyone. For this, we have performed user tests with our current drone. These user tests focused on the basis of a drone: flying. How comfortable are people when a drone is flying near them? We believe drones will become part of people’s daily lives. Important questions will arise, such as ‘what height and distance are appropriate?’, ‘What are the strengths of our current design?’ and ‘how is the drone perceived?’. Before the drone can be used in social situations, these questions need to be answered and improvements need to be made.
The drone: Blue Jay 1 with a dacron safety ring and a screen to display the eyes. Sonars and hardware are also visible.
Our drone does not look like every other drone as you can see in the picture. Our drone is a quadcopter of about 50 cm in diameter. The propellers are protected by the outer ring. The drone is still fairly loud, producing a noise of about 80 dB. Our drone also has eyes to make it look more friendly and less like a machine.
Setup of the experiment: the drone is flying towards the participants. The participant can tell the pilot to stop the drone at any moment.
How the drone is perceived and when people feel comfortable around it depends on the way it behaves. A good measure of how comfortable people feel around the drone is how close they allow the drone to fly. So, how did we find out how close the drone can fly to a person without them feeling uncomfortable or threatened? For this, we performed qualitative research with eleven participants. The drone took off at a distance of 11.5 meters from the participant. The drone was flying slowly but steadily towards the person at eye height, and he or she could tell the pilot to stop the drone at any moment. We measured this distance and asked the participant why they wanted the drone to stop at that distance. The drone then flew higher and lower and we asked the participants when they felt more at ease and safer. After this, we asked the participant about their feelings during the experiment. Additionally, we decided that the drone should not come closer than 150 cm to the participant for safety reasons.
The user tests gave us useful insights into the drone. First of all, 35% of the participants were confident enough that the drone worked perfectly and would have allowed the drone to come even closer than the recommended 150 cm. Most people, however, wanted the drone to stop earlier. When the drone flew too close to the participants they interpreted it as threatening and they no longer felt safe: “I felt like there was a chance that the drone might hit me”. Another useful finding was that most people experienced the sound of the drone as dissuasive. It is uncomfortable and disturbing: “It is the first thing you notice when the drone is flying”.
All but 2 out of the 11 participants thought the drone was most threatening when it was flying at eye level. When it flew lower people were more comfortable. One participant said “My legs are less important than my head”. Others said similar things or just that it felt less threatening. When it flew higher, the participants were also more comfortable. Most of them said this was simply because the drone flew further away and could fly over them.
What was mentioned more often than expected is that the outer ring of the drone gives a feeling of safety. People like the fact that the mechanical parts of the drone are hidden because the sight of the mechanics is what makes them feel uncomfortable; “The propellers are not visible at first, that is nice. I think if you hide all the hardware it would be better and give a better feeling about the drone”. This suggests that using a safety ring around the drone was a good design choice. Hiding all the hardware is a good improvement for the next drone. Besides the safe feeling, people also perceived the drone as interactive and some even perceived it as friendly. Most people said the eyes are “adorable” and make the drone “human-like”.
The user tests gave us the insight that we cannot use the drone in social situations yet. This is, however, what we want to achieve. At this moment, participants would see our drone fly as a deliverer of packages or in logistics. However, generally people are optimistic; “In the future, it is certainly possible to see a drone in the healthcare sector”. People obviously have to get used to the idea of a social drone. However, the current drone is not able to interact with people in a natural way. Changes have to be made in order to let the drone fly in situations where it can interact with people and assist them. To achieve this, Blue Jay needs to create a smaller drone, one that is quiet, able to fly stable, and never makes any mistakes.
Since the drone is supposed to become a social robot, it is important that it acts and looks friendly and is socially accepted. In matching behaviour and appearance, the function or goal of the robot should be taken into account. Research found that only if the human-likeness of the robot is related to the function of the robot, human-likeness was perceived to be necessary. People will find it more suitable that a social robot is human-like. Since the drone has to fly in social situations, it is important to make it look and feel human-like.
All in all, we found a lot of options for improvement of our next drones. A surprising and useful finding is that most people do not want the drone to fly at eye level. One of the most important improvements to be made is to reduce the noise of the drone. Besides that, we would like to cover the top and bottom of the drone to make it look safer. This is going to be a big challenge, but at Blue Jay we strive for the perfect drone. Next to the possible improvements, we also found some strengths. The eyes of our drone make it more interactive and user-friendly than any other normal drone. Also, our drone can fly reasonably close to people without them feeling uncomfortable. Of course, our drone needs to become able to fly autonomously. While developing the autonomy, we can take the results of these user tests into account. We can try to program it in such a way that the drone knows how high it should fly and what distance to keep. This is useful for the future because we believe drones will be part of society. They will be flying in all kinds of situations, even indoors. Therefore, we will keep performing these user tests and keep improving the drones for a future where people and drones can cooperate. We will create your future drone assistant!
Blue Jay on Edge!
From smart home devices to your social media feed, artificial intelligence is impacting our lives constantly. With Sophia becoming the first AI robot to receive an international visa, it is about time for everyone to have an AI assistant! Student team Blue Jay Eindhoven, with a vision to make your future drone assistant, is attempting to incorporate intelligence in its autonomous drones.
The team’s drone, which is named Blue Jay, has been built to look more friendly than the drones you might know from television or the internet. BlueJay already exhibits some human qualities like her eyes that respond to various situations, her autonomous (self-piloted) flight and her brand new ability of speech! As an example, last year she was able to assist lifeguards in swimming pools by being able to detect people drowning, followed by notifying the lifeguards and bringing a life jacket or first aid kit to the location.
For the upcoming months, together with SAS Institute, Blue Jay’s AI subteam is carrying out a Search and Rescue project. Using Deep Learning API and Event Stream Processing for Edge Computing software from SAS, Blue Jay will be able to accomplish a Search and Rescue mission with AI on edge. ‘On edge’ allows all the streaming data to be processed instantly for immediate response while also replacing the limitations of cloud based AI. Consequently, real-time detections will not display any delay leading to quicker actions from Blue Jay and the availability of internet connectivity will not limit her capability to react to any situation.
It is an exciting opportunity for Blue Jay Eindhoven to explore Artificial Intelligence of Things (AIoT) with SAS Institute. We invite you to follow us during the journey by keeping up with the upcoming blogs!
Basics of Technology #1
Ultra-Wideband positioning: How does the drone know where it is?
Welcome to the first Basics of Technology article! In this series, we show readers all the technologies that Blue Jay needs to function, and how they work. In this post, we will show you how Blue Jay uses Ultra-Wideband to determine her position. We have received the Ultra-Wideband system from Vention spin-off Focus.
By Thijs Jenneskens, Lead Software Engineer at Blue Jay Eindhoven
‘Wall-hugging’ prevention tech
So why do we use Ultra-Wideband? Well, we want Blue Jay to be able to avoid obstacles when she is flying somewhere. To do so safely, she needs to know her own location first. For example, we want her to know where the walls are in a certain building, so she can avoid a wall-hugging accident. If she does not even know her own location, this information is useless to her. Because of this, using Ultra-Wideband is, for us, the first step in creating the indoor drone of the future.
Fig 1: Blue Jay is happy to know she is not going to hug a wall anytime soon.
But what is it? Allow me to explain. Ultra-Wideband is a technology that enables both outdoor and indoor positioning. For example, you could think of a robot helping out in a warehouse. To be able to help move packages around in this warehouse, the robot has to have an idea of its position. This is where Ultra-Wideband shines. By using this technology, the robot is able to know where it is very precisely.
Here is how it works. The Ultra-Wideband location tracking system consists of three different types of devices. To achieve the goal of telling the robot its location, they work together very similarly to the way the navigation in your car works with GPS. GPS works because satellites know their own location, and are able to measure the distance to a car by measuring the time it takes to send a message to the receiver. Doing this with multiple satellites simultaneously, makes it possible to calculate the location of your car, using trilateration.
Fig 2: A car knows its location with some help of satellites.
In the Ultra-Wideband system the satellites are called ‘anchors’. These anchors first need to know their own location in order to be able to determine the location of what is tracked, the so called ‘tag’. Fortunately, we do not have to go through a time consuming process of measuring all the anchor distances ourselves because it is managed by the system. We put up the anchors around the space Blue Jay will be flying, and the anchors determine their own location relatively to each other. We also define the grid we want to use and automatically a map is created.
Blue Jay is equipped with not one, but two tags. This way we do not only know her exact location, but also which direction she is facing using the orientation of the two tags. The location is then shared with Blue Jay and with us, using the third device: The ‘access point’. The access point is the manager of the system and allows us to interface between the system and our computers.
The Ultra-Wideband system we use allows for countless possibilities. At Focus, a Vention spin-off, they developed the system specifically for use on stages for artists. You would not think of it, but having a spotlight follow artists on stage is harder than it seems. Often this is manually done or pre-programmed such that the artist has no freedom of movement anymore. By implementing the Focus system, the artists location is tracked automatically and translated to a pan, tilt, zoom and focus of the spotlight. Plugging this directly into the lighting engineers console results in automatic following the artist with your spotlight. This way light engineers and artists can focus on what they do best: Entertain their audience!
Lay3rs – The Ultimaker 2, 3D printer
While it’s not yet possible to fabricate electrical components like motors, sensors or even electronics using 3D printing, it’s entirely possible to 3D print drone parts like the frame, landing gear, and propellers. And these are usually the parts that break easily. Which makes 3D printing the perfect solution for them.
What is the benefit of 3D printed drone parts? If you’re not an experienced RC pilot, crashing and breaking parts of your aerial vehicle can be expensive. Instead of buying new parts, you can 3D print your own spare parts as needed. Thanks to our partner Lay3rs we have the opportunity to use our very own 3D printer. With our own 3D printer to hand, part designs can be easily upgraded and further developed to improve the performance of our drone. On top of that, 3D printed drone parts are quickly interchangeable, and we can do it ourselves. Building our own drone out of 3D printed parts serves as a great opportunity to learn about how it works and how it reacts to modifications we make.
Aside from vital components, there is a huge variety of extra things to add to our drone. We used the 3D printer for example to add more sensors to the system to make it more aware of its environment. When we wanted to test how different sensors compare when mounted on a drone it is easy to use 3D printed mounts in order to quickly interchange the sensor. Thereby more tests can be executed, and a better decision can be made. Even when we are not using the 3D printer for drone parts we use it for other useful tools like for example a business card holder.
One thing which makes our drone unique is the gripper underneath it. Since this is so uncommon there are no off the shelf grippers available on the market which meet the requirements we need. A 3D printer has made it possible for us to design and produce different grippers ourselves which are light enough to be suited for a drone.
We can conclude that a 3D printer is a must have for Blue Jay or any other project with the need to quickly iterate their product.
From care assistant to lifesaver: raising a drone with added social value
“I may be more experienced, but these young people come with fresh ideas. We learn from each other.”
Take a team of smart students from the Eindhoven University of Technology (TU/e) with a vision to develop an autonomous drone that adds real value in healthcare andrelated fields, make them aware of the ever-present threat of children drowning in swimming pools and you’ve got a challenge to get passionate about.
Despite the best efforts of thousands of well-trained lifesaving professionals and volunteers, on average each year 60 people will suffer injuries as a result of almost drowning in Dutch swimming pools, and 2 will actually drown. By far the most common group of victims is children.
According to Anja de Brouwer, a long-serving volunteer at the ‘Eindhovense Reddingsbrigade’ (the life rescue volunteers service in Eindhoven), who is trained to realistically simulate accident victims, “When a child’s drowning they usually don’t shout or wave their arms as you’d expect. It just goes quiet and, before you know it, you’ve lost them.”
Anja is one of the many volunteers who recently took part in a large-scale event at the Pieter van den Hoogenband Swimming Stadium organized by the rescue service to help the Blue Jay Eindhoven team develop a drone that can assist lifeguards by performing three key tasks: monitoring activity in a busy pool, detecting a swimmer in distress and responding appropriately.
During the evening, eight cameras continuously film the volunteers as they simulate drowning people. “To learn how to detect which movements mean someone may be in trouble rather than playing, the drone must be shown lots of different videos of drowning people,” explains Liza Boormans, the Blue Jay team manager. “Once it has enough data, it can recognize and detect drowning behavior itself.” Thankfully, film footage of people really drowning is rare. Hence the need to film the LOTUS volunteers role-playing victims.
Not waving but drowning
This process of detecting someone drowning is one of the real innovations of the lifeguard assistant project, and one of its biggest challenges. “There are already detection methods on the market that spot if someone has drowned, but the Blue Jay team is developing a drone to detect if someone is about to drown — in time to help them. That’s a much more complex challenge,” says Arjen van de Wetering, an Internet of Things (IoT) consultant at IBM who has been supporting the Blue Jay team since IBM became a main partner to the project a few months back.
“Distinguishing between play and distress is difficult using traditional analytics methods” explains Arjen. “So the team is using artificial intelligence (AI), employing IBM’s Watson technology. The team will use the same drowning simulation videos to train the system and to test whether it can make the drowning-playing distinction, and later whether it can make that judgement effectively in real-life situations.”
This detection process isn’t simple. Having singled out individuals from its aerial view of the whole pool, the system must follow people using approximately 2 seconds of consecutive video frames to determine how someone is moving. The oldest frame from those 2 seconds is then dropped and replaced by a new one. The system is thus continuously analyzing how people move within a 2-second timeframe. The calculations to determine whether a swimmer’s movements constitute playing or drowning take a huge amount of computer power. In our first pilot setup, all these videos are streamed live from the cameras to the cloud, where the calculations are then done on IBM’s Watson AI mainframes. The ‘decision’ about the swimmer’s movements is then made by Watson and sent back in a matter of milliseconds to the drone, which acts accordingly — a classic IoT solution. For the final system, the team probably don’t want to depend on an unreliable internet connection, so it might consist of a trained Watson-based application on-premise. But that decision has yet to be taken.
International and multi-disciplined
Blue Jay Eindhoven is an on-going project with the ultimate goal of developing a drone that can assist professionals in healthcare settings. The team is made up of students from the Technical University of Eindhoven (TU/e) who take time out from their studies to join Blue Jay for a year. “Because you get no elective points or financial support for working on Blue Jay,” explains Liza, “only people who are very committed to our goals and vision join the team.” Liza’s own role as team manager ends in September when she returns to her full-time studies in Medical Sciences & Technology.
The current team is an international group of 18 students from 9 different academic disciplines. As the Blue Jay personnel change pretty well entirely each academic year, continuity is assured through an advisory board that includes some of the previous year’s students who support the next year’s team. “The board ensures the team follow the long-term Blue Jay vision,” says Liza, “but gives them sufficient freedom to add their own insights and innovations.”
Last year’s team focused on developing a drone to assist in the Máxima Medical Center near Eindhoven. Much progress was made. It could, for example, play games with child patients. But a stumbling block, which has not yet been overcome, is noise: a drone is still pretty noisy and that obviously isn’t handy in a hospital environment.
The perfect setting
Ironic, then, that it was during a brainstorming session on solving the noise problem that the team heard about the case of a 9-year-old refugee girl who drowned in a swimming pool in Rhenen. They immediately saw the potential for a Blue Jay drone to help prevent such tragedies. “Swimming pools are a perfect setting,” explains Liza, “Noise isn’t an issue, but we can still demonstrate the huge potential value of the drone: the ideal proof of concept.”
But as team project manager, Mike van Sighem, points out, the swimming pool drone has its own challenges. “As well as drowning detection, it must be able to autonomously navigate buildings, notify personnel, and bring bulky and heavy objects to an accident scene.”
Pim van Dommelen, Blue Jay’s mechanical engineer, sees this last as probably the biggest mechanical challenge of the lifeguard drone. “It needs to be light, otherwise you have to switch batteries every few minutes, and a drone that’s out of the air too often isn’t much use as a lifeguard assistant. But at the same time, it must be powerful enough to fetch heavy objects such a lifebuoy, defibrillator or first aid kit.”
Partners make perfect
NXP, the second main project partner, is helping the team with making the drone both robust and responsive enough to do its job. “Last year we helped the Blue Jay team develop a robust communication channel between ground station and drone,” explains Gino Knubben, Principal System Engineer at NXP and closely involved with the company’s support to Blue Jay. “This year they’re using a slightly newer version of the same technology, but our role is even more drone-focused, looking mainly at two major systems. First, the flight control system, which deals with the drone’s movement, orientation, balance and remote control – the team chose this year to switch to our flight controller, NXPhlite. And secondly, the application processor, which deals with the display and the interaction.”
For the project, functionality is critical and one big benefit of working with world-class companies like NXP, IBM and Fourtress, the third and final main Blue Jay sponsor, is the Blue Jay team’s access to advanced technology not found in consumer products. “Compared to regular Wi-Fi on your phone,” explains Gino, “our automotive-qualified connectivity solutions are superior in robustness, RF reflections, power consumption and the number of ad-hoc connections you can do per second.” The robust connections come from 802.11p technology used in the automotive industry, and can handle the drone’s rapid movement and ad-hoc communications. Gino laughs, “Two years ago, everybody had to remember to switch off the Wi-Fi on their phones before testing or demonstrating. Now with 802.11p we’ve resolved that particular interference issue.”
Another benefit of having NXP as a partner is that, unlike many semiconductor manufacturers, they can provide a wide range of components. “We’re providing a large part of the electronics they need for drone development,” says Gino. “Not just the wireless link but also high-performance microcontrollers; proven reference boards with all the components assembled; and even complete sub-systems, including supported Operating Systems. The team then adds their peripherals, displays, sensors, applications, graphical user interfaces, etc.”
But Gino is first to say the collaboration is not one-way traffic. “The Blue Jay team are basically beta-testing some of this technology for us and their feedback is invaluable. For example, on development of the FMU (Flight Management Unit): along with all the other partners involved, they provide feedback, they test… It’s outside the students’ comfort zone and the remit of the project. So that wider commitment is great. They also visit us at NXP on a regular basis, which is a lot of fun.”
And there are areas where old hands learn from the students. “They’re very knowledgeable, for example, on the ‘total drone’, such as how to calibrate and configure it. And have some seriously skilled drone pilots! Another example is hardware assembly experience. We’d been assembling drones at NXP ourselves, which took a lot of time. Then Jari van Ewijk, electrical engineer from Blue Jay turned up and built one in under 90 minutes. I asked how he could do it so quickly. He said ‘how many prototypes you think I’ve had to build? We’ve crashed quite of few of them…’ He could probably have done it blindfold! The next day we received pictures and movie clips of their first flight with that NXP-drone kit!”
Another spin-off from Blue Jay is that the technology involved can have applications beyond the narrow (albeit very important) fields of lifesaving and hospitals. “For example, there’s huge potential for FMU technology and drones in areas like warehouse logistics,” says Gino. “So the beta-testing also contributes to making this a proven solution NXP can offer in other contexts and sectors where the same parameters apply.”
Fourtress provide software solutions for technical systems. “We’ve been involved since the first Blue Jay team,” says field manager, Douwe Gerritsen. “Initially, we simply provided technical knowledge. But as we have engineers with pretty strong drone-related experience, for the past two years instead of money we have loaned the team an engineer for 3 months.”
It’s win-win as the engineer also gets interesting experience. He has worked on the low-level systems, Lidar, obstacle detection, the drone’s inner workings, the ultra-wide band and data communication systems. Fourtress also provided a student on placement with them to work on the Blue Jay mobile app development and drone-API as their thesis project.
“Probably our biggest challenge in developing the app is that Blue Jay team projects are very dynamic,” says Douwe. “They set ambitious goals. That’s inspiring, of course, but means implementation tends to change along the way as they encounter problems. You try to work with that flow, so you don’t become an obstacle yourself. But sometimes you need to help them set boundaries in terms of what’s achievable within the timescale.
“There’s also the technical challenge of trying to fly a drone in populated environments, including the paradox of wanting a drone that both works totally autonomously, but is also controlled by people.” This means a person must be able to use the app to give the drone commands, after which the drone figures out for itself how to execute the task. Sometimes deciding autonomously, for example, which route to take or task to prioritize. But the interface needs to let people easily set new tasks and get a clear overview that updates them on what each drone is doing at that moment.
“The app is basically the drone’s manager. It gives the drone tasks, but the drone decides how to execute them. The prioritization aspect is interesting. It’s important the user can schedule tasks to be done later or immediately. But also, that they get feedback on how quickly the drone will actually do the task. Because when an emergency happens it will override everything else. This variety of tasks makes things a lot more complex.”
Another challenge is notifications. “The tricky part is that people will use their phones not just to control the drone, but also to surf the internet, take calls, etc. Technically, this means the app must be able to run a service in the background even when it’s not active, so it can still give users notifications, for example of an emergency or a completed task, when their drone app isn’t active.
Douwe sees the Blue Jay partnership as an invaluable one for Fourtress. “How we operate with Blue Jay is a bit of a model for how we try to operate with all our customers. In that, if you hire one Fourtress person, you hire all of Fourtress: all our knowledge is at your disposal.”
Blue Jay gets students out of their ivory tower and working with real-world organisations to address real-world needs. So it’s reassuring when society responds so positively. “The Reddingsbrigade were very enthusiastic from the start,” recalls Mike. “It’s cool how they’re so open to adopting new technology. Ultimately, we share the same goals. By working together, we can help each other achieve those goals better and quicker.”
At IBM, Arjen has quickly become a huge fan of the students and what they’re doing. “Like much of my own work, this project is at the forefront of technology. I bring my experience to the project, but these young people have a different perspective and come with fresh ideas, so we definitely learn from each other.” Douwe echoes this view on behalf of Fourtress. “Because with Blue Jay you’re working with such enthusiastic and dedicated people, it makes it so much fun.”
At NXP, Gino is particularly impressed by how the Blue Jay team leverage maximum value from partners’ expertise and function so maturely as a unit. “They take technology from us for the drone, from IBM for in the cloud, use other partners for the mechanical chassis… operating just like an efficient little company, each with their own areas of responsibility. Of 18 team members, I only deal with 2 or 3 hardware engineering, electronics and software people. For young, inexperienced students it’s pretty impressive.”
With her hands-on experience, LOTUS’ Anja is equally excited by the drone’s potential. “It’s basically an extra pair of eyes. Even the best lifeguard gets tired, but not the drone. It stays alert, and that means once it’s fully operational it will help ensure people are safer.”
Those days are still a way off. For now, it’s all about proof of concept. Even so, the team will continue to work on a drone that can provide first aid in emergency situations. Because, as Arjen says, “if one day the drone saves just one life, the whole project will have been worth it.” And you don’t need AI to detect the truth in that statement.
get in touch with the team if you’re interested in supporting their work, getting involved or becoming a partner. Send an email to: email@example.com
PCB Design 2.0
Earlier on, we wrote about the design of our own PCB. It has now been printed and is ready to be integrated and tested on the drone! You can see the result in the picture below.
Four new drones
We’re expanding our drone family by four new drones! This is keeping us very busy, since we have to order lots of extra components and then assemble everyting together step by step and integrate the electronic components onto the frame. Are you curious how the assembly process looks like? We made a little time lapse below, where you can admire how a new Blue Jay is created from scratch.
Currently, our drone is covered with wires connecting all the different components. However, this will soon be over, since we are now designing our very own PCB, on which multiple components will be integrated together. This will not only save a lot of space, but also prevent wire ruptures. The picture below shows the design of the board. With the help of Bestronics, we will get it printed in the next weeks and we hope to integrate it on our drones before the final demonstration in June.
Blog Holland Innovative – TU/e Student Start-ups
Zoals een andere studenten start-up in een eerder schreef: “Op de Technische Universiteit Eindhoven hebben ze studententeams. Dit zijn groepen studenten die een groot doel hebben, …”
Zo is Blue Jay Eindhoven ook één van de studententeams. Wij ontwikkelen als start-up een product dat uniek, creatief én innovatief is, namelijk: indoor, autonome en intelligente drones voor in de gezondheidszorg. Het multidisciplinaire team (lees negen verschillende studieachtergronden) bestaat uit negentien studenten afkomstig van over de hele wereld.
“… zoals meedoen aan een landelijke of zelfs wereldwijde wedstrijd met het product dat ze zelf hebben ontwikkeld.”
Maar stel je nou eens voor dat er (nog) géén wedstrijd, race of competitie de right fit is voor jouw innovatie. Is er dan alsnog voldoende uitdaging om een studententeam te vormen? Jààzeker!
Team Blue Jay ontwikkelt een maatschappelijk relevant product. Vliegen van A naar B in 0,1 seconde sneller dan een ander team brengt ons niet dichter bij ons doel. Echter, een drone die van A naar B kan vliegen, daarbij zowel objecten als mensen kan vermijden en ondertussen nog “ff” bij drie patiënten de hartslag kan meten, dat klinkt al een stuk interessanter, niet? Wij spreken van een overwinning als er een nieuwe functionaliteit is toegevoegd aan de drone (én het werkt); kampioen worden staat op de planning!
Juist ja, een planning…Plannen. Cruciaal. Maar hoe krijg je bij een zeer diverse, dynamische en eigenwijze groep studenten de neuzen dezelfde kant op? Nou, laten we beginnen bij het begin.
Afgelopen zomer vond de overgang plaats van team Blue Jay#2 naar Blue Jay#3. Vanuit het oude team werd het idee van een drone in de gezondheidszorg meegegeven, maar een concrete visie ontbrak. Vanaf het begin was duidelijk dat het team weinig ervaring had in projectmanagement en dat de opstartfase (te) langzaam ging. Mijn team had behoefte aan een coach die enthousiast werd van het idee om ons actief te begeleiden bij projectmanagement, on the job te willen coachen, ons kon uitdagen en ons vooral niet zou overladen met theorieën en termen waar het world-wide-web al vol mee staat.
Klinkt allemaal mooi en aardig, maar ja, waar vind je dan zo’n enthousiaste, flexibele en motiverende “projectmanager coach” die bereid is om teams, bestaande uit studenten notabene, naar een next level te brengen? Mijn motto: iets dat moeilijk is, is niet onmogelijk. Zo kwam ik in contact met Roel Wessels van Holland Innovative.
Duikbootgedrag is contact met de markt uitstellen
Tijdens onze eerste werksessie met Roel maakten wij kennis met de term “Duikbootgedrag”. We hadden Roel het plan voorgeschoteld dat we aan het eind van het academisch jaar een spectaculaire demonstratie wilden geven voor onze partners en stakeholders. Echter onze visie was onduidelijk en de einddemonstratie had noch vorm noch invulling. Maar goed, met het idee dat we nog één heel jaar voor de boeg hadden probeerden we onszelf gerust te stellen. Technische kwaliteit? Ja, dat had het team zeker. Hoge motivatie? Dubbelcheck! Het zou allemaal wel goed komen. Een perfect voorbeeld van Duikbootgedrag. En ik kan je vertellen, daar kom je bij Roel niet mee weg.
Roel maakte ons ervan bewust, dat er nieuwsgierigheid gewekt kan worden door de buitenwereld mee te nemen in het ontwikkelproces. Iets tonen aan het publiek dat nog niet af is, is juist mega interessant en zorgt voor vroege feedback. Zo kwam al snel het idee om halverwege het jaar een Droneshow te organiseren voor onze partners en stakeholders. We beseften maar al te goed dat alleen een presentatie geven over de vooruitgang van Blue Jay niet voldoende en overtuigend genoeg was. Want: een drone die niet kan vliegen, is natuurlijk geen drone. Nee, de drone zelf moest kunnen tonen wat er al bereikt was. We legden ons als team dan ook een harde deadline op: midden februari 2018, dan moest en zou ie vliegen. Autonoom.
Een idee zonder uitvoering is een No-Go
Tijdens onze sessies met Roel gingen we aan de slag met product breakdowns. Daar gebruikten we geen fancy computerprogrammaatje voor, maar grote papieren vellen en stiften. Aan de hand van deze breakdowns bloeiden er heftige discussies op, kwamen er taken aan het licht en ontstonden tussen de verschillende disciplines in het team nieuwe samenwerkingsverbanden. Als je dit zo leest zou je bijna zeggen, zoals Roel dat op z’n Brabants zou doen, dat gi van ‘n lèije dèkske. Verre van! Er kwam namelijk duidelijk naar voren dat een heldere visie voor Blue Jay ontbrak en dat we deze nodig hadden als basis voor zowel ons eigen team, maar ook voor de teams die in onze voetsporen zullen treden.
Dit zette het team vroegtijdig op scherp. Haarscherp. De knop ging om. Er werden meetings ingepland met bestuursleden, zorgverleners en patiënten van ziekenhuizen en zorginstellingen. Parallel werd er aan een technische roadmap getimmerd met als gevolg dat in rap tempo drie verschillende drone prototypes de lucht in schoten. Stukje bij beetje werd er een visie gevormd waar zowel de user, partners als het team warm van werden. Twee maanden voor de demonstratie gingen de uitnodigen al de deur uit met de aankondiging van “Blue Jay’s Droneshow” en zoals je weet #bijzijnismeemaken: men keek er naar uit en zou hiervoor speciaal naar Eindhoven komen! Dat hadden we toch mooi voor elkaar gekregen. Eén dingetje: de drone stond toen meer op de grond dan dat ie in de lucht te vinden was…
Het publiek voelde de team spirit
Hoe smooth alles in de planning ook op elkaar is aangesloten, in praktijk overkomen je altijd dingen die je niet had voorzien. In de periode rondom een evenement wordt het altijd extra duidelijk dat een studententeam, met alle respect, geen kantoorbaan van 9-tot-5 is. Van projectmanagement tot aan cable-management, van warme bitterballen voor de gasten tot aan het lettertype van de presentatie: het moest en zou kloppen.
En hoe mooi is het dan, als het op het moment suprême gaat zoals zowaar gepland?! Plus dat het publiek de chemie binnen het team en de team effort zowel kon zien als voelen. Een échte Team Prestatie! Maximaal Genieten. Overwinningsgevoel.
Toekomstig spektakel door een gebalanceerde roadmap
Maar (Roel vindt vaak dat ik teveel “Maar”) hoe krijg je het team dan weer scherp terwijl het overwinningsgevoel overheerst? Een overwinning betekent een stap dichter bij het kampioenschap, maar er zijn nog een aantal belangrijke wedstrijden te gaan waar we ook moeten vlammen…
Zo staat er voor ons momenteel nog het eindevenement gepland voor midden juni. Hét evenement om te laten zien wat Team Blue Jay in één jaar heeft weten te bereiken. Sinds januari zitten we om de tafel met verschillende partijen om tezamen te kijken naar de invulling van dit evenement. Aan de hand hiervan zijn requirements opgesteld en de laatste weken zijn we druk geweest om hierin prioriteiten te stellen en afwegingen te maken. Want wat voor de voortgang van Blue Jay cruciaal is, levert niet altijd direct het “wow-effect” op bij de buitenwereld. En andersom vragen sommige gewilde functies te lange ontwikkeltijd.
Toch hebben we dit, als team, weten te vertalen naar een technische roadmap die moet leiden tot een spectaculaire onthulling in juni en die stiekem meer zal zijn dan alleen een onthulling. Namelijk een grote stap dichterbij onze droom: een drone assistent voor in de gezondheidszorg.