ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
Interview questions###Interviewee 1 (Andrey)Interviewee 2 (Ilya)Interviewee 3 (Dmitry)
2
Tell us, what are you doing in motorsport or developing car systems now?I am currently developing a drift car for participation in winter drift competitions.
I'm currently involved in rallying, but I'm not developing car systems.I participate in track days, I drift for myself and at local events.
3
Do you work with drift? Is it more for a show, training, or competition?For me, it's more about the show and the atmosphere of the competition itself.Yes, drift is one of the important components in a rally. It is used not for show, but as a competitive element: drift helps to pass turns faster, that is, at a higher speed.More for practice and pleasure, sometimes for competition.
4
What technologies or telemetry do you use in drift analysis?I use data from the car's ECU to track parameters during a drift. Special telemetry sensors are also provided at competitions to track the trajectory.Mostly none. The rally uses very light cars, and they don't have extra sensors on them, so as not to add weight. There is a struggle for every kilogram, and unnecessary technology gets in the way.Only basic: GPS log, action camera, sometimes OBD.
5
Have you heard about autonomous drift projects (for example, from Toyota Research, Stanford, etc.)? How do you feel about them?Yes, I've heard of them, but I'm skeptical about them.No, I haven't heard. I am neutral.Yes, I've seen projects from Toyota and Stanford. It looks impressive, but it's far from reality for us.
6
What is the most difficult thing for you in controlling a drift in practice?The most difficult thing is the ability to smoothly shift in a turn.The most difficult thing is to keep the car skidding along the correct trajectory. Skidding is easy, but keeping it is difficult. Especially because the coverage is often uneven, and you need to feel it.Control the car in the transitions and catch the moment of loss of grip in time.
7
Have you ever lost a car in an uncontrolled skid? Why was this happening?This is due to the fact that the rear wheels rotate too fast and do not have time to catch on the coating.Yes, I have. Most often, this is due to a sudden change in the clutch: the hook appears, then disappears. It takes very precise operation of the accelerator pedal to maintain the desired dynamics.Yes, most often it's a late reaction or an incorrect assessment of the clutch.
8
Is there a way to understand in advance that the car is about to go into a skid?Yes, you can understand, this happens when the engine speed rises sharply.Yes. For example, if the engine picks up or slows down, you can predict that the clutch is changing. Also, by the position of the body and the sensations, you can understand that the car is losing or, conversely, "catching" a hook.Sometimes it feels like the steering wheel, vibrations, sound. But that's subjective.
9
What parameters do you usually monitor in order to “feel” the car in drift (angles, speed, gas, clutch, etc.)?I focus mainly on personal feelings.Usually it the trajectory of the car, the angle of the car, the direction in which it "looks", and the engine speed.Gas, steering angle, speed, grip and first of all, the behavior of the rear axle.
10
If there was a tool that says, "in 0.5 seconds, skidding will begin," would that be useful? In what way exactly?For a person involved in drifting, it would be useless, because drifting is about feelings and personal control.Such a system is more useful for ordinary drivers or trucks. Similar systems already exist: ESP, DSC, and so on. For rally or drift, it is unlikely.Yes, especially for beginners. Even a simple display would help you navigate.
11
Do you think AI/algorithms can be useful in drift training or error analysis?Yes, it could be useful for analyzing the trajectory and improving it in the future.I don't think so. Such techniques require practice, not theoretical hints. It's like in a drift school: first the instructor shows, and then the student repeats and learns through mistakes. Text or advice is no substitute for physical experience.Definitely. Error analysis after check it is very useful.
12
Would you trust a car to drift on its own? Under what conditions?This would allow me to study my race, the car would reproduce my route passage, and I could disassemble it and find mistakes.No. I don't trust autopilot. Unless it's at a closed landfill, in a car that doesn't belong to you, and in complete safety, then maybe for the sake of experiment.Not yet, but it's interesting to see. In a safe environment, yes.
13
f it were possible to connect a module that predicts skidding and controls the machine in these conditions, would you test it? Why/why not?I think not, it is impossible to predict skidding, skidding must be controlled by the driver.No, I believe that skidding cannot be predicted, it must be controlled by a person.Yes, if you can turn it on/off. I would like to compare the behavior.
14
What would be an indicator of the success of such an algorithm for you (accuracy, stability, security, etc.)?An indicator of success would be the stability of the races.If the system ensures that we don't just crash into something, that's good.Stability, security, and clear logic.
15
Would you like to participate in testing or see how it works in a simulation?Yes, as a participant in such races, it would be interesting for me to test how realistically the system can To participate is not. But it would be interesting to see how it looks and works.Yes, especially if you can look at simulations or an RC model.
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100