Driving Simulator For Applications In Human Adaptive Mechatronics Engineering Essay

Published:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Nowadays, human machine system considers being a proven technology, which has gained an important role in various human activities. One of the most recent developments in this field is Human Adaptive Mechatronics (HAM) approach for enhancing human skills. This approach therefore is different compared to an ordinary man-machine system, in terms of its ability to adapt to the changes in its surroundings and in the human changing level of skills. The main problem in HAM however is in evaluating the human skills level on a machine manipulation. Therefore, this paper deals with the proposed formula to quantify and classify the skill index of human operator. The methods to normalize time and error also discuss later in this paper. An experiment involved a car driving simulator is carried out, to justify the proposed formula in a machine manipulation system. As stated in the formula, results show that the time and error are both important in determining the skill of human operator.

1. Introduction

The human machine system (HMS) is relatively synonymous in human life, where most of the machines need humans to operate. Any machine that involves human can be categorized in HMS, even though it is an open loop system, such as washing machine, which only requires human to press a few buttons on it to work. However, this system requires a human to have in-depth understanding of the machine operation, which is a one-way relationship [1, 2]. In other words, only the human is required to learn the machine and its aspects.

The existence of a new improvement system called Human Adaptive Mechatronics (HAM), gives hope that human and machine can have a two-way relationship, where the machine is able to understand the human, provide feedbacks and responds to the human action. HAM is defined as intelligent mechanical systems that adapt to the human skills in the different situations, assist to improve the human skills and assist the human machine system to achieve the best performance [2, 3, 4]. For example, when a human subject is driving a vehicle, the current technology requires that only the human has to fully understand the vehicle from various aspects such as handling, safety and reliability. If a HAM system is implemented on the vehicle, it should be able to understand the human, not only know how it was handled, but also recognize the human skill. In HAM, skill is defined as the ability to perform a task with fast speed, less error, high repeatability and good problem solving strategy [3, 4]. By understanding the human skills, the machine can distinguish between different humans. Then, by using these differences, the machine can provide feedbacks or responds based on the human skill level. In other words, if the human is high skilled, then the HAM machine only gives minimum assistance if needed compared to low skilled human, where the machine will provide maximum assistance in whatever conditions. However, until now, there is no specific method to quantify the skill in realization of a HAM system [5, 6].

This paper proposes a formula to quantify the skill index of a human operator based on time and error. In literature, there are also other researchers who have proposed method for skill classification, but it can be proven wrong, complicated or limited to specific application [7-12]. For example, the study from Sasaki [8] provides the formula to measure the skill index based on time and error, but the formula is proved to give a human low skill index when the experiment time is fastest and the error is smallest. With that combination, the formula should consider the human as high skilled [4].

2. Proposed Skill Index Formula

2.1 Formula development The proposed formula is based on Table 1, which evaluated using the logical conditions and the definition of skill in HAM. Noted that En is normalized error, Tn is normalized time and J is skill index. For example, if a human operator executed a task in a fast time (F = Fast), and with a small percent of error (S = Small), the skill index, J of the human operator is rated as very high skilled (VHS). Similarly, if the execution time is medium (M), and the percentage of error is large (L), then the skill index, J given as Low Skilled (LS).

In developing the formula, every term of error and time on Table 1 has been replaced with an equivalent maximum value, such as Small (S) = Fast (F) = 0, Medium (M) = 0.5, Large (L) = Slow (Sl) = 1. Similarly, the terms of J are also replaced with the equivalent values, i.e. VHS = 1.00, HS = 0.75, MS = 0.50, LS = 0.25 and VLS = 0.00. These values are determined between zero and one that divided into eight segments as shown in Fig. 1. The values in box are selected for each J term, so that the range is 0.25 between each term. Hence, Table 2 shows the complete skill values used for each combination of time and error.

HS

VHS

LS

MS

VLS

1.000

0.750

0.500

0.000

0.875

0.625

0.375

0.250

0.125

Fig. 1 Selected value and range for each skill level.

Table 1 Combination of normalized error, En and normalized time, Tn for each skill index, J.

En

Tn

J

S

F

VHS

S

M

HS

S

Sl

MS

M

F

HS

M

M

MS

M

Sl

LS

L

F

MS

L

M

LS

L

Sl

VLS

Legend:

S - Small, M - Medium, L - Large,

F - Fast, Sl - Slow, VHS - Very High Skilled,

HS - High Skilled, MS - Middle Skilled, LS - Low Skilled, VLS - Very Low Skilled.

Table 2 The equivalent values used for En, Tn and J.

En

Tn

J

0.00

0.00

1.00

0.00

0.50

0.75

0.00

1.00

0.50

0.50

0.00

0.75

0.50

0.50

0.50

0.50

1.00

0.25

1.00

0.00

0.50

1.00

0.50

0.25

1.00

1.00

0.00

Based on Table 2, the formula to quantify the skill index is developed as follows. Let say the linear relationship between J and (Tn, En) is shown in equation (1).

J = ATn + BEn + C (1)

where

A, B, C = constants.

By substituting the value of En, Tn and J from Table 2, the values of A, B and C can be computed and are obtained as 1, -0.5 and -0.5, respectively. Therefore, the final formula for J is described in equation (2).

J = 1 - 0.5(En + Tn) (2)

The range of J, En and Tn are from 0 to 1.

2.2 Normalized time In any experiment, if the actual execution time is use, the range of time is between zero and infinity. This gives difficulties in measuring the skill of the human operator due to large range. Therefore, the normalized time is used so that the range is only between zero and one. Formula to normalize time is shown in equation (3).

(3)

where

TB= the best theoretical time by assuming the track is a straight line, ignoring the corners and braking, and using the maximum speed.

t = time achieved by each subject.

Based on equation (3), a human subject can obtained zero in normalized time if t = TB, which is the fastest time. Similarly, he/she can obtained one in normalized time when the time is the slowest (t = ∞).

2.3 Normalized error The range for actual error is also between zero and infinity. Therefore, an error is normalized using equation (4) to have range between zero and one.

(4)

where

Es = the smallest possible error that obtained by the sampled subjects in specific application in the nearest 2 significant figures, and Es < e.

e = actual error obtained by each subject.

Based on equation (4), if a human subject obtained the smallest error, then En = 0. But, for larger error, En is near to one.

Table 3 Range for each level of J and assist from a machine.

Skill Level

Range

Level of assist

VLS

0.000 ≤J ≤0.125

Full

LS

0.125 <J ≤0.375

High

MS

0.375 <J ≤0.625

Medium

HS

0.625 <J ≤0.875

Low

VHS

0.875 <J ≤1.000

None

2.4 Classification of Skill The proposed skill index divides into five levels, that is VLS, LS, MS, HS, and VHS. Based on Fig. 1, the range for each level is determined. The best level is VHS, ranges between 0.875 and 1. The worst level is VLS, consists of 0 to 0.125. Table 3 shows the complete range of each skill level and the proposed level of assist from a machine. For example, if a human subject is very low skilled (VLS), then the machine required to give full assistance in order to achieve the best performance.

3. Experiment setup

Based on literature, many researchers using a computer simulator to identify the skill of human operator in experiment because the actual HAM system is not yet established [8, 13, 14, 15, 16]. Hence, an experiment using a driving simulator is carried out. Moreover, driving a car is considered as one of the complex mechanical tasks that involved skill, human and machine [16]. In experiment, 15 human subjects are selected from different driving backgrounds. But, in this paper, only data from two selected subjects are used for comparison and justification of the proposed formula. The hardware setup for experiment is shown in Fig. 2.

15-inch monitor

Steering wheel

Keyboard

Pedals

Fig. 2 The hardware setup for the HAM skill index experiment.

3.1 Driving simulator Fig.3 shows the graphical user interface (GUI) of the driving simulator used in the experiment. This driving simulator is designed using Visual C++, and capable of recording the time and the responses of the human subjects. The simulator can be handled using either a Logitech Momo steering wheel with pedals or a keyboard. The maximum speed in the car simulation program, Vmax = 600 unit/s.

Fig.3 GUI of designed driving simulator.

3.2 Error Measurement From experiment, the tracking error is evaluated as the difference between the track, T and the driving path, P created by the subject. Equation (5) shows the formula to measure the error.

(5)

where

e = average tracking error (obtained by human)

D = minimum distance for every point of T and P.

Np = number of points for driving path.

3.3 Experiment features In order to get different skill indices from every subject, five different tracks were used, i.e. straight, circle, ellipse, square and triangle. For each track, five trials are carried out during the experiment. This is to identify the operational skill of each subject by doing repetition. Detail features of every track used in experiment is explained in Table 4.

Table 4 Features of track used.

Track

Shape

Distance (Unit)

TB

(s)

Things to measure

Expected skill

Straight

5000

8.3

Skill to follow linear and continuous line.

HS

Circle

9500

15.8

Skill to follow nonlinear line and continuous turning.

HS

Ellipse

11000

18.3

Skill to follow other type of nonlinear line and continuous turning.

HS

Square

13500

22.5

Skill to follow linear line and 90oof turnings.

MS

Triangle

10000

16.7

Skill to follow linear line and 60oof turnings.

MS

4. Results

Fig. 4 shows graphs for normalized time and error for straight track over five trials. Although both subjects obtained almost similar time but their normalized errors are different. For first trial, subject A has large error but it dramatically improves for other trials. On the other hand, subject E improved his error after two trials.

Fig. 5, 6, 7 and 8 show the results for circle, ellipse, square and triangle tracks, respectively. For circle track, after five trials, both subjects obtained almost same error. The time to complete the track is also same for both subjects.

For ellipse track, the completed time for both subjects remains the same again. Although subject A obtained the smallest error for trial three but it increased for last two trials. For trial five, both subjects obtained the same normalized error. For square and triangle tracks, although subject A is slower, but his error is smaller and better than subject E. Normally, when the time is fast then the error is large and vice versa.

Fig. 4 Normalized time and error for straight track.

Fig. 5 Normalized time and error for circle track.

Fig. 6 Normalized time and error for ellipse track.

5. Analysis

Based on proposed formula and data from Fig. 4, 5, 6, 7 and 8, the skill index is quantified for each subject. Table 5 shows the operational skill for Subject A and E within five trials. It seems that skill index for subject A improved after two trials in three tracks, i.e. straight, circle and triangle. For subject E, his skill index achieved 'high' after three trials on two tracks, i.e. straight and triangle. In last three trials, both subjects have improved their operational skill.

Due to both subjects obtained almost similar execution time for straight, circle and ellipse tracks, the skill index largely depends on normalized error. If the normalized error is small, then the skill index is high. The normalized time only gives little effect for these three tracks.

Fig. 7 Normalized time and error for square track.

Fig. 8 Normalized time and error for triangle track.

For square and triangle tracks, both normalized time and error have important roles in the quantification of skill index. For example, for triangle track, at trial two, although subject A is slower, but his skill index is 'high' because his error is smaller. At the same point, although subject E is faster, but his error is large which gives him a 'middle' skill index. Overall, the quantification of skill index is affected by either the time or error and the different environments.

By knowing the skill of human, it is easy for the HAM machine to determine the level of assistance required. Therefore, Table 5 also provides the proposed level of assistance based on overall skill index. Generally, it seems that both subjects are high skilled for all tracks even for square and triangle tracks, that consist of sharp corners. This shows that both subjects require only low level of assistance from the HAM machine. Although it is low, but among five tracks, both subjects need more assistance in following the triangle track.

Table 5 Operational skills with values for Subject A and E.

Subject

Track

Skill index

Average Skill

Level of assist required

Trial

1

2

3

4

5

A

Straight

MS (0.490)

HS (0.856)

HS

(0.832)

HS (0.858)

HS (0.863)

HS (0.780)

Low

(0.220)

Circle

MS (0.491)

HS (0.778)

HS

(0.801)

HS (0.829)

HS (0.778)

HS (0.735)

Low (0.265)

Ellipse

HS (0.666)

HS (0.761)

HS

(0.855)

HS (0.831)

HS (0.759)

HS (0.774)

Low (0.226)

Square

HS (0.744)

HS (0.821)

HS

(0.726)

HS (0.866)

HS (0.677)

HS (0.767)

Low (0.233)

Triangle

MS (0.526)

HS

(0.662)

HS

(0.662)

HS (0.665)

HS (0.757)

HS (0.654)

Low (0.346)

E

Straight

MS (0.522)

MS (0.538)

HS

(0.860)

HS (0.865)

HS (0.858)

HS (0.729)

Low (0.271)

Circle

HS (0.650)

HS (0.735)

HS

(0.683)

HS (0.763)

HS (0.774)

HS (0.721)

Low (0.279)

Ellipse

HS (0.798)

HS (0.739)

HS

(0.692)

HS (0.710)

HS (0.743)

HS (0.736)

Low (0.264)

Square

HS (0.733)

HS (0.743)

HS

(0.764)

HS (0.764)

HS (0.722)

HS (0.745)

Low (0.255)

Triangle

MS (0.556)

MS (0.602)

HS

(0.656)

HS (0.683)

HS (0.771)

HS (0.654)

Low (0.346)

6. Conclusions

A new formula to quantify the skill of human operator for HAM application has been proposed. In order to justify the formula in machine manipulation system, an experiment using a driving simulator that controlled by a steering wheel with pedals is carried out.

The formula's ability to quantify and classify the human skill is justified in various environments and a number of trials. It confirms that both time and error have important role in determining the human skill.

For future work, the proposed formula will be used to quantify the skill in sudden transitory conditions when driving a car, such as tire puncture, slippery and accident. In these situations, a human really needs assistance from the machine in order to overcome the problems.

Writing Services

Essay Writing
Service

Find out how the very best essay writing service can help you accomplish more and achieve higher marks today.

Assignment Writing Service

From complicated assignments to tricky tasks, our experts can tackle virtually any question thrown at them.

Dissertation Writing Service

A dissertation (also known as a thesis or research project) is probably the most important piece of work for any student! From full dissertations to individual chapters, we’re on hand to support you.

Coursework Writing Service

Our expert qualified writers can help you get your coursework right first time, every time.

Dissertation Proposal Service

The first step to completing a dissertation is to create a proposal that talks about what you wish to do. Our experts can design suitable methodologies - perfect to help you get started with a dissertation.

Report Writing
Service

Reports for any audience. Perfectly structured, professionally written, and tailored to suit your exact requirements.

Essay Skeleton Answer Service

If you’re just looking for some help to get started on an essay, our outline service provides you with a perfect essay plan.

Marking & Proofreading Service

Not sure if your work is hitting the mark? Struggling to get feedback from your lecturer? Our premium marking service was created just for you - get the feedback you deserve now.

Exam Revision
Service

Exams can be one of the most stressful experiences you’ll ever have! Revision is key, and we’re here to help. With custom created revision notes and exam answers, you’ll never feel underprepared again.