Takashi Shinoda

  • Engineer

Takashi Shinoda

Winning 3rd prize for Human powered airplane which I designed in Tokyo Institute of Technologies. Studied optimization of CFRP structure at master course. After graduate, joined Mitsubishi Heavy Industry and participated for Boeing 787 composite wing structure analysis which was under development. I also have experienced the first manufacturing test for the composite wing structure.
Next, moved to a company making flying bike and made the simulator and control logic for it.
By sympathizing the vison of ‘Automating Any Machine’ I have joined DeepX from july 2021.


When I was a university student, Genetic algorism was at the top of hype cycle. At that time, I had chance to know about neural network which will become deep learning for nowadays. The approach of simulating the human nerves felt like fantasy and very different approach from deterministic approach of mechanical engineering. I was not able to image the neural network get the big progress.

Was very curious about flying mobilities. So, after graduating master course I got a job for designing an aircraft made by CFRP which was the latest technology. I was imagining that designing an aircraft is done by state-of-the-art technologies with fancy tools like FEM and CFD. However, proving the safety was much more steady effort. The culture was ‘analysis supported by test’. Every, analysis should be validated by test before it can be used for design. What an impressing way it works.

I also had change to participate for the first product manufacturing test. Every day even if you are making same thing with same process, new problem came up. Even if there are slight difference with the material, ware of machine, the product should be at same level of quality. I didn’t realize the difficulty until I experienced. The difficulty compared to designing was deferent type.

Things seems easy may be very difficult and things seems difficult may be very easy. For example, ‘flying a rocket is just mixing hydrogen and oxygen and burn it.’ Is something like not knowing the complexity. Human cannot realize it’s real difficulty until he/she do it by them self. Some work looks easy may be very difficult for machines to do it. Especially things which need decisions are not easy for traditional machines.

Recently I was developing a drone control algorism. Drone and robot seems very deferent. However, it has same behavior doing [recognition]-[planning]-[actuator control] repeatedly and is categorize to same field robotics. The ecosystem for robotics has grown much. Using the experience for mechanical design, manufacturing and robotics, I am excited how much I can challenge to the human being ability.



  • Control system for automating machines
  • Physics simulation environment for test and develop automation algorism
  • Recognition of environment which is necessary for intelligent machine
  • Challenging to the world by startup
  • Exceed the ability of machine by human manual control


  • I like all kind of sports. During I was a student I was playing table tennis and tennis. After starting work, I joined a club activity for bicycle, joining 8 hours endurance race at Suzuka circuit as team. I also like hill climb race which you can ride at your speed. The Fuji hill climb is a good rice to try for beginners.
    After I moved to Tokyo, there is not much opportunity to ride bicycle. To avoid my body getting hard, I go to gym several times a week.

  • I totally had no skills of playing instruments. However, I was curious of playing it. Started learning guitar just before I get 30 years old. I attended a group lessen but I was not able to catch-up with the progress. So, changed to personal lesson. I also tried with the band session. I was extremely hard time for me, because I could not adjust to the lyrism. I kept continuing practice because it is my hobby. After 5 years, now I think I am getting used to it. It seems nothing is impossible if you keep challenging.

We use cookies on this site to enhance your user experience. If you continue to browse, you accept the use of cookies policy on our site.