Yoon Dong-Guk is an artificial intelligence researcher who creates software using artificial intelligence and shares open source so that everyone can utilize these technologies and devices. Recently, he is developing a wheelchair capable of autonomous driving. What does he mean by “independent and universal AI”?

Q You’re researching artificial intelligence wheelchairs capable of autonomous driving.

A While contemplating how to realistically consume artificial intelligence related to autonomous vehicles and the rest of the technologies that have been under study, the ‘autonomous driving electric wheelchair’ was designed. After grasping the basic limitations of commercially available electric wheelchairs, research was initiated by limiting the use of the frame of a folding manual wheelchair as it is. We paid attention to essential functions such as in-wheel motor method, torque that can overcome at least 15% inclination, driving distance of 80km assuming 10km/h speed, and convenient battery removal. This research is being conducted as a member of Z-LAB.

Q I am curious about the results of the research so far.

A Currently, it has been developed up to the first stage, where partial assistance driving is possible. Based on the American Society of Automotive Engineers, the first stage of autonomous driving is a level in which the system partially intervenes, such as maintaining speed and distance between cars, and maintaining lanes. At the beginning of the study, the goal was set to stage 5, but it was judged that stage 5, fully autonomous driving, was not plausible in the current situation, and the goal was reset to stage 3 autonomous driving. In stage 3, the system is the driving control body, and the system also plays a role in detecting variables during driving. Now that the first stage has been achieved, we plan to develop it from 2021 to the second and third stages. The project is expected to be finalized by 2023.

Q In what ways can wheelchairs become autonomously driven?

A Control logic uses ‘Kinetics’ and ‘Kinematics Model’ based on the dynamics of ‘2 Wheel Base’. We are developing a control that uses 10 Axis AHRS to acquire Local coordinates through wheelchair acceleration, angular velocity, and progression information, and combine them with control logic. Additionally, using VRS/RTK GPS, global coordinates can be obtained in outdoor spaces. The target positional accuracy is within 2cm error, and is currently within 5cm error. For vision, use ‘LiDAR/w Camera’. The object is detected from the image entered through the camera, and the distance value according to the detected object is measured with the Mono-Depth neural model. This enables automatic emergency braking (AEB) and avoidance maneuvers. Wheels used in self-driving wheelchairs were originally made for electric bicycles and were fused through good will beyond their original use. Thanks to this, the prototype was completed in five months.

Q People with disabilities will be able to enjoy freedom with this wheelchair

A With this project as an opportunity, I wanted to take a deeper look into the issue of “equal mobility rights”. The “Personal Navigator,” which helps visually impaired people freely board the bus, was created in the same context. The electric wheelchair currently used is quite heavy enough to reach 100kg, and it is not free to move due to its four wheel base. In addition, there are statistics that 1 in 10 people have an accident. The newly developed wheelchair not only has an autonomous driving function, but also is light. The current prototype is 28kg, and it is expected to be lowered to 18kg in mass production. In the future, it is expected that users of self-driving artificial intelligence wheelchairs can move freely and safely anywhere. On the other hand, it is expected that it will be able to help people with disabilities in their economic activities beyond “equal mobility rights”. As an example, because the pulling force is strong, economic activity will be possible if it is used for the last stage of the parcel delivery system, the “last mile”.

Q Will our perceptions change as a result?

A Not long ago, I went to Jeju Island with a prototype of an autonomous electric wheelchair. Airline services for wheelchair users were well-equipped, but we recognized that non-disabled people were very violently dealing with the people with disabilities. In particular, I experienced firsthand that caring was lacking when driving at the airport for a wheelchair test. “If a world is comfortable for people with disabilities, it is a world where everyone is comfortable” This is the truth. We all need a change of perception. Changes in perception through autonomous driving artificial intelligence wheelchairs are a goal to be achieved. In addition, I hope that someday, the autonomous driving artificial intelligence wheelchair will reach a time when it is evaluated solely as a ‘personal mobility’. And I believe it will.

Q What kind of company is WizGene?

A It started as a study lab where people who research artificial intelligence in each field gathered, and it was established with the goal of contributing to society with the technologies they have. WizGene is always thinking about how the vague technology of artificial intelligence can be applied to reality. This is because there are many cases where the use is vague or only plausible and not easily applied to the present reality. WizGene not only solves social problems through research and technology development on artificial intelligence, but also pursues artificial intelligence as an’ independent and universal technology’.

Q What does it mean to give back technology to society?

A WizGene will continue to share its research and development technologies and devices as an ‘open source company’. Software engineers are familiar with this ecological structure. WizGene’s members, who are positioned as seniors in their respective fields related to artificial intelligence, take this good donation as their philosophy. That’s why it is described as ‘universal and independent AI’. I hope there will be a future where anyone can utilize these technologies and devices.

Q As science and technology advance, another kind of inequality or disparity society is expected.

A For example, in the case of wheelchairs, ‘equal mobility rights’ should be observed. However, not everyone can use the same wheelchair; this is the logic. Although standards must be adhered to, I think that it is an inevitable social phenomenon that not everyone enjoys the same things equally. However, it is natural that efforts by all of us, including the government, are needed to close the gap. WizGene is preparing to establish an Institutional Review Board (IRB), a bioethics committee that specializes in artificial intelligence to prevent various ethical problems caused by the development of science and technology. It is an institution that can judge whether there is any deviation in securing bioethics and safety, and when conducting artificial intelligence research on humans, it must be approved by this committee.