- Email: [email protected]
Evaluation of an intelligent seat system
Applied
WORTH EINEMANN
Ergonomics
Vol26, No 2, pp. 109-116, 1995 Elswier Science Ltd Printed in Great Britain. ooo3-6670/95 $10.00 + 0.00
0003-6870(95)00006-2
Evaluation of an intelligent seat system David Ng, Tom Cassar and Clifford M. Gross Biomechanics
Corporation of America, Melville, NY
11747, USA
This study was conducted to evaluate the effect of an intelligent seat system, a microprocessorbased interactive seat that automatically adjusts itself to fit a seated individual by making pressure-sensitive adjustments on its own. Fist, a standard American automobile seat (‘baseline’ seat) was assessed for comfort. Subjective ratings of comfort, pressure distribution and seated anthropometric measurements were recorded for 20 test subjects. These measurements were recorded while the subjects maintained a simulated driving position in a seat buck. The comfort scale was based on a rating of 1 to 10, with 1 corresponding to ‘very poor’ and 10 corresponding to ‘very good.’ Based on a nonlinear, multiple regression model that had been previously developed, the comfort rating of the seat was predicted based on the subjective ratings and the recorded values of 450 pressure measurements from 20 subjects. The predicted comfort value was 7.46 for the baseline seat. Following the baseline assessment, the intelligent seat system was installed into the standard American automobile seat. The objective and subjective assessments were then repeated for 17 subjects and the new predicted comfort rating was 8.06. A t-test performed on the subjective and objective measures indicated that this was a significant improvement in seat comfort. Overall, subjects felt the self-adjusting seat was more customized and more comfortable, providing a better fit. Keywords: seat comfort, pressure distribution,
biomechanics
Prolonged sitting during daily driving activities can develop stress in muscles of the back, buttocks, and legs. Well-designed seating support can reduce this stress. When the body is not well supported, several muscle groups act together to restore stability, contributing to static loading. As a result, the occupant feels discomfort and fatigued. A well-designed seat should be able to accommodate all sizes and shapes of users and should provide adequate support (Gross et al, 1992). In this paper, we relate seat comfort to a biomechanical variable - pressure distribution between the seat surface and its occupant - and then describe enhancement of the comfort of the occupant by automatically adjusting the amount of support provided by the seat. Biomechanical research conducted on over 1100 seat-subject evaluations, comparing load distributions with subjective ratings, showed that comfort ratings can be predicted from patterns of weight distribution (Gross et al, 1994). This information allowed construction of multiple regression equations that predict comfort levels from seat pressure variables, leading to the development of the Intelligent Seat System (Figure I), a microprocessor-based interactive device that senses the pressures at the body-seat interface using sensors placed under the upholstery. Based on the occupant’s pressure distribution, the
AIR CEL
PRESSURE SENSING DEVICE
IIMICRO-+I I’
I-T-T
CUTOFF t!
II
PROCESSOR CONTROLLED II I
IIswTTcHMc IIL
Figure 1 Intelligent seat system schematic. High-pressure system, pressure measurement sensors, automated power inflate-power deflate on all bladders to optimize comfort, microprocessor-controlled switching. US Patent Number (1) 5,060,174 (2) 5,170,364 (3) 5,283,735
109
110
D. Ng et al
system decides which adjustments to make and automatically makes those adjustments by inflating embedded air bladders to optimize comfort. As a result, support and pressure distribution are improved. Over time, if the user changes posture or position, the Intelligent Seat System re-adjusts itself accordingly. A good distribution minimizes load concentrations that can restrict blood flow and affect the nerves, causing discomfort or pain. For example, pressure should not be concentrated primarily at the ischial tuberosities, but rather be distributed across the buttocks and thigh areas for better support. Sanders and McCormick (1987) suggest that the weight should be distributed rather evenly throughout the buttocks area, but minimized under the thighs. The objective of the study reported here was to evaluate the effect of the Intelligent Seat System after it was installed into a standard American automobile seat (called the ‘baseline’ seat). A comfort assessment was performed on the baseline seat without and then with the Intelligent System. Comparisons were made to see if there was a significant difference in seat comfort.
Table 1 Importance of seat features: percentage giving response (ranking from most important seat features to least important seat features) Ratings” Features
1
2
1. 2. 3. 4. 5.
65 70 35 70 60
3.5 30 60 30 20
Seatback recline Lumbar support Height of seatpan (from Seatpan tilt Head rest position
floor)
3 0 0 ;;20
“Ratings are based on the following descriptions: 1 - It is very important to me that this adjustability feature is provided 2 - This adjustability feature is not important to me, but I feel it should still be provided 3 - I do not care for this adjustability feature
Table 2 Importance of seat attributes: percentage giving response (ranking from most important attributes to least important seat attributes) Ratings’
Methods Subjects
The subject group consisted of 20 healthy individuals (10 male and 10 female), who were college students, skilled workers, and professionals. Subjects were independently recruited and hired. The average age of the subjects was 28.7 years, the average weight was 71.5 kg and the average height was 179.9 cm. The age of the subjects ranged from 19 to 51 years, with a mean of 28.7 years. The subjects’ driving experience level ranged from 3 to 35 years, with a mean of 12.1 years. A questionnaire was administered to the subjects to determine the important features of a car seat. Tables I and 2 summarize the subjects’ perception of the importance of seat adjustability features and seat attributes. As shown, 70% of the subjects felt that lumbar support and seatpan tilt are very important, while only 35% felt the height of the seatpan is very important. Also reported as very important were the seatback recline and headrest position. Subjects indicated that their perception of seat comfort was influenced most by thigh support (75%), thoracic support (70%), lumbar support (65%) and presence of armrests (65%). The seat was mounted on a seat buck, which simulated the driver’s side of a car (Figure 2). The seat buck was equipped with a steering wheel, a ‘dead’ pedal mounted at an angle of 40” for the left foot, and a simulated accelerator pedal (mounted at 60”) for the right foot. The baseline seat was mounted so that the front mounting bolt was 22.5 in (572 mm) from the heel point (accelerator pedal). The seat buck also had a clearance of 11.5 in (292 mm), which represented the height of the floorboard above the ground.
Features
1
2
I. Lumbar support 2. Seatback firmness 3. Texture and material of the upholstery 4. Seat cushion firmness 5 Thoracic support h: Presence of armrests 7. Seatback size 8. Buttocks support 9. Seatback lateral support 10. Thigh support 11. Physical appearance of the seat 12. Seat cushion size 13. Seat cushion lateral support 14. Head/neck support 15. Colour of the seat
65 50
35 45
0
2.5 60 70 65 50 60 15 75 50 IO 15 55 30
40 30 25 25 30 30 40 20 40 35 40 45 SO
35 10
3
IO
20 IO 10 JO
55 45 20
“Ratings are based on the following descriptions: 1 - This attribute of the seat greatly affects my perception of overall seat comfort 2 - This attribute of the seat somewhat influences my perception of overall seat comfort 3 - My perception of overall seat comfort is not at all affected by this attribute
Seat pressure measurement
Seat pressure measurements were obtained using thin flexible pressure mats (Figure 3) made up of forcesensing resistors. These pressure mats were configured over the seatpan and seatback and did not affect seat
Figure
2
Seat mounted on a seat buck simulating the driver’s
side of a car
Evaluation of an intelligent seat system
geometry. Two pressure mat configurations were used, one for the seatpan and one for the seatback. The seatpan mat was designed to cover an area of approximately 16 X 16 in (407 X 407 mm), which included 225 pressure points arranged into a 15 X 15 matrix. The seatback mat measured 14 X 21 in (356 X 533 mm) and also consisted of 225 pressure sensors arranged into a 15 X 15 matrix. At zero or very low pressures, the sensors performed as an open circuit. After the pressure had reached a low threshold, increasing force rapidly reduced electrical resistance. Data were collected from the seatpan and seatback pressure mats simultaneously with the subject seated at his or her preferred setting. During data collection, the subjects assumed a driving posture with both hands gripping the steering wheel, the right foot on the accelerator pedal and the left foot on the ‘dead’ pedal. Seat pressure data for the seatpan and seatback for each subject were obtained for the baseline and intelligent seat at a frequency of 45 Hz for 15 s. Subjects did not move throughout the entire interval. The averages of the seat pressure data for the seatpan and seatback were divided into 12 regions (Figure 4) for analysis. For each region, the mean and standard deviation of pressures, the sum of pressures, and the maximum and minimum pressure gradients were calculated.
111
‘.
-
Figure 3 Pressure mat consisting of 225 force-sensing resistors
Anthropometry
Sixeen seated anthropometry measures were recorded from each participant. All measurements were made using an anthropometer and a goniometer. These measurements were taken to ascertain the seated posture of the test subjects. Table 3 lists the ranges of acceptable body angles for a seated driver (Babbs, 1979). These figures are presented for comparison with the actual values reported in the results section.
1 9 1
Subjective measures of comfort
Data collection procedure
Each subject was briefed on the methodology and purpose of the experiment and then given a short questionnaire to determine level of driving experience and familiarity with different car seats. Subjects were also asked to rate generally the importance of certain seat adjustability features and attributes (as listed in Table 4). Subjects familiarized themselves with the seat and then adjusted it to their preferred seatings. After
6
7 6
5
Two questionnaires were designed to obtain subjective comfort ratings for various features of the seatback and seatpan. The first questionnaire required the subject to rate comfort on a scale of 1 to 10, with 1 representing ‘very poor’ and 10 representing ‘very good.’ The second questionnaire had a scale of +lO to -10 representing each extreme condition of the factor, with the midpoint 0 representing ‘just right.’ For example, for seat cushion firmness, -10 corresponds to ‘too firm’ and +lO corresponds to ‘too soft.’ These comfort ratings were based on a continuous interval with a normal distribution, as in previous studies we found it was consistent and mathematically more convenient to assume a normal distribution.
10
[
Seat Pan
L
1 Seat
Seat pan
Seat back
I 2 3 4 5 6
7 = R&t
= = = = = =
hght bolster Left bolster Right buttock Let? buttock hght thlgb Left thigh
Back
bolster
8 = Left bolster 9 = Right thora‘x
10=LeftthOraClC I I =Right lumbar I2 = Let? lumbar
Figure 4 12 regions of seat pan and seat back
Table 3 Recommended
rangesa for body angles for a seated driver
Angle
Range (degrees)
1. 2. 3. 4. 5.
la-45 m-120 9.5-120 95-135 90-110
Shoulder angle Elbow angle Hip angle Knee angle Ankle angle
“Babbs (1979)
D. Ng et al
112
achieving a feel for the seat, subjects were asked to rate the ‘comfort’ of each seat feature based on a scale of 1 to 10 (Table 5). Following the initial comfort rating, pressure data were acquired for the seat at the subject’s preferred setting and subjects were then given the second questionnaire with a rating scale of -10 to +lO for the features in Table 6. Seated anthropometric dimensions were then measured for each subject at the preferred settings. Adjustment ranges, physical measurements and characteristics of the baseline seat were also recorded (Tuble 4).
After the evaluation of the baseline seat, the Intelligent Seat System was installed. Air bladders were placed in strategic areas of the seatpan and seatback. A microprocessor-controlled switching device operated a reversible pump to inflate/deflate the bladders. After
Table 5 Mean ratings of comfort for the baseline and intelligent seat (with standard deviations). Scale: 1 = very poor to 10 = very good
Comfort
responses
1. Seatpan cushion
Table 4 Baseline seat characteristics
Light grey Leather surface Yes Yes 12 Electric Electric Electric Manual None Yes Manual
Colour Trim/texture Bolsters on pan Bolsters on back Number of adjustments Fore and aft control Back recline control Seat height control Lumbar adjustment Thigh bolsters control Adjustable seatpan Headrest Range of adjustment/dimensions Fore-aft (cm) (from gas pedal to front of seatpan) Min Max Seatback recline angle (degrees) (from horizontal) Min Max Seatpan height (cm) Min Max Height of lumbar curvature apex (cm) (above seatpan) Thickness of lumbar curvature (max) (cm) (at apex) Seatpan tilt (degrees) (from horizontal) Seatback with (lumbar area) (cm) (inside bolsters) Seatpan width (front) (cm) inside bolsters)
2 Seatpan width 3 Seatpan depth 4. Seatpan angle 5 Thigh support 6. Buttocks support 7. Overall seatpan support 8. Backrest cushion 9. Backrest height
25.4 48.3
10. Backrest width 11. Backrest recline
104” 150”
12. Lumbar support 13. Centre height/apex of lumbar support
27.9 33.0
14. Upper back support 19.1 15. Overall backrest support 3.91 16. Ease of accessing controls 10” 17. Ease operating controls 53.6 18. Overall comfort 29.9
Baseline (N = 20)
Intelligent (N = 17)
7.8 (0.98) 7.8 (1.28) 7.4 (1.20) 8.0 (1.46) 6.8 (1.32) 7.6 (1.06) 7.8 (1.02) 8.4 (1.08) 7.6 (1.48) 8.0 (1.20) 8.0 (1.34) 8.2 (1.24)
9.0 &92)
(1.48) 8.6 (0.86) 8.8 (1.16) 9.0 (0.90)
7.6 (1.36) 7.8 (1.42) 7.8 (1.18) 9.0 (1.34) 9.0 (1.46) 8.2 (1.04)
9.0 (0.86) 8.4 (1.70) 9.0 (1.06) 9.4 (0.80) 9.4 (0.88) 9.2 (0.76)
(1.42) 8.8 (0.88) 8.8 (0.84) 9.0 (0.94) 9.0 (0.90) 9.0 (0.94) 8.8 ;;08)
Table 6 Percentages of comfort ratings for the baseline seat and intelligent seat Feature/attribute
Baseline scat
Intelligent seat
1. Seatpan cushion firmness 2. Seatback cushion firmness
Too firm 20 20
Just right 55 55
Too soft 25 25
Too firm 11.8 29.4
Just right 76.5 70.6
Too soft 11.8 0
3. Seatpan height
Too high 10
Just right 70
Too low 20
Too high 5.9
Just right 88.2
Too low 5.9
4. Seatback size
Too large 0
Just right 75
Too small 25
Too large 5.9
Just right 76.5
Too small 17.7
5. Seatpan tilt
Too forward 10
Just right 85
Too backward 5
Too forward 0
Just right 94.1
Too backward 5.9
6. Seatback recline angle
Too reclined 10
Just right 85
Too upright 5
Too reclined 0
Just right 94.1
Too upright 5.9
Too much 10 20 10 0 1.5
Just right 40 60 35 30 40 55
Too little 50 20 55 70 45 45
Too much 17.7 5.9 29.1 5.9 0 0
Just right 82.3 88.2 58.8 35.3 88.2 82.4
Too little 0 5.9 11.8 58.8 11.8 17.7
7. 8. 9. 10. 11. 12.
Pressure on thigh area Lumbar support Upper back support Shoulder support Seatpan bolster support Seatback bolster support
0
Evaluation of an intelligent seat system
installation of the hardware and software was complete, 17 subjects from the first comfort study was recalled (3 of the 20 were unavailable) for a follow-up comfort assessment of the baseline seat with the intelligent technology incorporated into it. Seat geometry, cushion and configuration were unaffected. Subjects adjusted the seat to their preferred settings and then pressed a button to start the intelligent system. The system sensed the subject’s pressure distribution at the body-seat interface. Based on the pressure distribution, the seat then self-adjusted by inflating bladders in various areas (eg lumbar, side bolster) of the seat to provide more support, optimizing user comfort. After the seat had self-adjusted (about 35 s), ten samples of pressure data were collected and two subjective comfort questionnaires were administered.
113
on the intelligent seat. Figures 8 and 10 (seatpan pressure distribution) illustrate that the pressure was well distributed across the buttocks (55.6%) and thighs (37.4%). Figures 6 and 12 (seat back pressure distribution) display pressure as being most concentrated in the three regions: (lower, middle and upper lumbar) of the lumber area (76.5%). The rest of the pressure was distributed throughout the thoracic region and side bolsters,
Thorecic 16.24%
LUInbar bolsters 7.05%
Thoracic bolsters 2.30%
Results Pressure distribution
The load distributions for the seatback and seatpan regions are summarized in Figures 5-8. These values are expressed as a ratio of the pressure on an individual region to the total pressure on the entire seatpan or seatback and are used to determine whether proper support is provided to the occupant. Pressure plots were obtained by averaging the pressure distributions of all the subjects per seat and calculating the pressure gradient at each of the different points, Plots of the seatpan and seatback regions were made separately with contour intervals of 5% (data normalized to maximum values for the seat-subject combination, which was then assigned a value of 100%). Figures 9 and 10 show the average pressure gradient distribution for the seat pan and seat back for 20 subjects on the baseline seat. It can be seen from Figures 7 and 9 (seatpan pressure distribution) that pressure was not uniform across the buttocks and thighs. Most of the load was concentrated in the buttock region (73.9%), while 18.5% was found in the thigh region. Figures 5 and IO (seatback pressure distribution) show pressure to be most concentrated at the lumbar region (55.9%) and thoracic region (43.2%). Little pressure was found in the thoracic bolsters (0.2%) and lumbar bolsters (0.7%). Figures I1 and 22 show the average pressure gradient distribution for the seatpan and seatback for 17 subjects
Lumber 76.48%
Figure 6 Intelligent seat back pressure distribution
Pan Bolsters
Thighs 184%
Buttocks 73.64%
Figure
7
Baseline seat pan pressure distribution
Pan Bolsters Thoracic 4X22%-
bo~e~~~,%
Lumbar bolsters 0.68% Lumber 56.86%
Figure 5 Baseline seat back pressure distribution
Buttocks 56.66% Figure
8
Intelligent
seat pan pressure distribution
D. Ng et al
114 Subjective responses
The means of the subjective comfort responses before and after installation of the Intelligent System are listed in Table 5. The baseline seat had high ratings for ease of accessing and operating controls (9.0), backrest cushion (8.4), lumbar support (8.2), seatpan angle and backrest recline (8.0), and relatively low rating for thigh support (6.8). After installation of the Intelligent System, subjects’ perception of comfort increased substantially; for example, thigh support increased from 6.8 to 9.0, overall backrest support increased from 7.8 to 9.0 and lumbar support increased from 8.2 to 9.0. Table 6 shows rating of seat features and attributes from the second subjective comfort questionnaire for the baseline seat and intelligent seat. The percentage of subjects who rated ‘just right’ for pressure on the thigh
SEAT
PAN
PRESSURE
area increased from 40% of 20 subjects (baseline assessment) to 82.4% of 17 subjects (intelligent seat assessment). An increase in the ‘just right’ rating for the intelligent seat was also shown for the following features/attributes: lumbar support (60-88.2%), upper back support (40-58.8%), seatpan bolster support (4088.2%), and seatback bolster support (55-82.45%). Seatpan height, seatpan tilt and seatback recline angle were each rated as relatively high, possibly owing to subjects adjusting the seat to their preferred settings. Predicted comfort
A comfort model was used to predict seat comfort based on pressure distribution (Gross, 1992; Gross et al, 1994). The model was a multiple regression function that expressed the relationship between the
DISTRIBUTION 36
SEAT PAN PRESSURE DISTRIBUTION 0 2 4 6 7 9 11131517192022242628303133353739414344
33
g
33
30
30
27
27
24
24
$21
21
ln 18
18
x
15
w '5 0 m 12 z .P cc
orl’li J.l/r-J ’ ’ ’ ’ ’ ’ ” ’ I’IAiYi o 2 4 6 7 9 11131517192022242628303133353739414344° Front
of
Seat
SEAT 0
BACK
PRESSURE
6
6
3
3
0 0
- 27
27 i Y
- 24
:: 24 m
29 .g 6
10
Y
Seat
Pan
2
BACK
PRESSURE
DISTRIBUTION
4
6
7
9 11131517192022242628303133353739414344
4
6
7
9 1113151719202224262830313335
27
- 21
5
- 18
Ln"18
- 15
‘i; 15 Q,
21
- 12
w
-9
z9 .E"
-6
L1:6
12
Bottom
Back
Average pressure gradient distribution of all subjects for the baseline seat back
of
:: 24 m
9 11131517192022242628303133353739414344*
Seot
9 11131517192022242628303133353739414344°
SEAT
- 30
of
7
11 Average pressure gradient distribution of all subjects for the intelligent seat pan
0
133
Bottom
6
DISTRIBUTION
30
7
4
Figure
2 4 6 7 9 11131517192022242628303133353739414344 IIIIII//IIII
6
2
Front
33
Figure
g
”
9 Average pressure gradient distribution of all subjects for the baseline seat pan
36
12 9
Pa7
Figure
36
Figure
12
of
Seat
Bock
Average pressure gradient distribution of all subjects for the intelligent seat back
115
Evaluation of an intelligent seat system
dependent variable (comfort) and 17 independent variables (pressure). The final model was derived as a result of establishing the closest or best-fit relationship by using a combination of independent variables and eliminating those that contributed least to the comfort model. The pressure variables were obtained from the baseline seat and the intelligent seat. Pressure variables from previous car seat assessments were also included in the model to strengthen the correlation of comfort and patterns of pressure distribution. Based on a scale from 1 to 10, with 1 being ‘very poor’ and 10 being ‘very good’, and based on the values of pressure variables from the other seats, the predicted comfort value was obtained. The perceived comfort value reported here is an average of the 18 comfort responses provided by the test subjects from all 18 questions administered in the first questionnaire (Table 5). The predicted and perceived comfort values are presented in Figure 13; the comfort values for the intelligent seat (predicted = 8.06, perceived = 8.9) were higher than for the baseline seat (predicted =7.46, perceived = 7.9).
Table 7 Mean seated anthropometric measures (with standard deviations) for the baseline seat Measurements’
Mean
Sitting height
103.6
2. Elbow height
(4.6) 59.2
1.
3. Left knee height (2.9) 41.9
4. Right knee height 5. Left buttock-popliteal 6. Right buttock-popliteal
length length
(9.6) 86.7
8. Right ankle angle
(6.0) 134.9
9. Left knee angle knee angle
%.? it&;)
11. Left hip angle (6.1) 99.9
12. Right hip angle Subject anthropometry
Seated anthropometric data on all subjects are summarized in Table 7. These measurements were taken after the seat was adjusted to the preferred settings for each subject. The mean anthropometric values were compared with the recommended range of body angles listed in Table 3 (Babbs, 1979). Shoulder and elbow angles were found to be above the recommended ranges, possibly owing to seated postures in which subjects stretched their arms out more to reach the steering wheel. Hip, knee and ankle angles were found to fall approximately within the recommended ranges.
(3.1) 46.5 (2.3) 105.8
7. Left ankle angle
10. Right
(4.8) 48.8
13. Left shoulder 14. Right shoulder
(6.7) 49.0 (13.5) 52.1
angle angle
15. Left elbow angle
%2
16. Right elbow angle
(24.6) “All linear measurements measurements in degrees
in centimetres
and all angular
Conclusions Statistical analysis
A t-test was performed to determine whether a significant difference in comfort based on subjective results (Table 5) and objective results (predicted values) existed between the baseline seat and the intelligent seat. With a null hypothesis of no difference in comfort between the two seats, the results showed a significant difference in comfort (p = 0.001); that is, subjects felt significantly more comfortable after the intelligent system was installed.
12 10 8 6
4 2 0
Piedicted comfort
Perceived comfort
Figure 13 The predicted and perceived baseline and seat and intelligence seat
comfort
for the
The baseline seat had good ratings for ease of accessing and operating its controls, seatback cushion, lumbar support, seatpan angle and seatback recline. It was not well rated for thigh support and shoulder support. Most subjects liked the seat and lumbar adjustability features, and they found the lumbar support to be adequate. However, for most subjects (taller ones in particular) there was not much contact under the thighs, and they felt that more thigh support should be provided as well as a deeper seatpan depth. Subjects felt that the seatback and pan bolsters were inadequate, and that the bolsters were hardly noticeable. Even though the seatpan and back cushions had good ratings, most subjects felt that the cushions should be firmer. They also felt more lateral support should be provided in the seatpan and upper back. Pressure distribution data showed an uneven distribution across the thighs and buttocks, with most of the load concentrated in the buttock region. There was good pressure distribution in the lumbar and thoracic regions. Consistent with the subjective responses, loads on the thoracic and lumbar bolster regions were low, which indicates that more support should be provided in these areas. The comfort ratings for the intelligent seat were higher in almost all cases. Thigh support, upper back support, lumbar support, side and pan bolsters received a much better response compared to the baseline
116
D. Ng et al
assessment. Subjects rated the intelligent seat as more customized and accommodating, and providing a better fit. They particularly liked the way the seat adjusted automatically. Pressure distribution data showed better distribution across the buttocks and thighs compared with the baseline seat. In the seat back, pressure was mostly found in the lumbar area. This could be attributed to the fact that the three lumbar regions (lower, middle and upper) have provided greater support, relieving most of the pressure in the upper back. Common methods to distribute pressure in seat design are the use of contouring and seat cushion firmness. These produce a fixed shape, which does not provide easy adjustment to occupants of different sizes and weights, and as a result occupants may feel discomfort or fatigue during prolonged periods of sitting. An intelligent seat can avoid this by adjusting automatically to occupants of different sizes and weights, redistributing pressure and providing more support, thus optimizing user comfort and safety. This study concluded that subjects felt significantly more comfortable in the baseline seat after the Intelligent System was installed.
Acknowledgements The authors wish to thank the staff of McCord Winn Textron for their participation and support with the Intelligent Seat System, and acknowledge the excellent microelectronics design contributions for the Intelligent Seat System by Mr Jeff Finkelstein, of Microprocessor Designs, Shelburne, Vermont.
References Babbs, F.W. 1979 ‘A design layout method for relating seating to the occupant and vehicle’ Ergonomics 22, 227-234 Gross, C.M. 1992 ‘A biomechanical approach to ergonomic product design’ in Maynards Industrial Handbook 4th edn, MacGraw-Hill, New York, pp 8.45-8.61 Gross, C.M., Goonetilfeke, R.S. Menon, K.K., Banaag, J.C.N. and Nair, C.M. 1992 ‘Biomechanical assessment and prediction of seat comfort’ Automot Technol Int pp 329-334 Gross, C.M., Goonetilleke, R.S., Menon, K.K., Banaag, J.C.N. and Nab, C.M. 1994 ‘The biomechanical assessment and prediction of seat comfort’ in Hard Facts and Soft Machines: The Ergonomics of Seating Taylor & Francis, London, 1994 Sanders, M.S. and McCormick, E.J. 1987 Human Engineering Design McGraw-Hill, New York
Factors
in
WORTH EINEMANN
Ergonomics
Vol26, No 2, pp. 109-116, 1995 Elswier Science Ltd Printed in Great Britain. ooo3-6670/95 $10.00 + 0.00
0003-6870(95)00006-2
Evaluation of an intelligent seat system David Ng, Tom Cassar and Clifford M. Gross Biomechanics
Corporation of America, Melville, NY
11747, USA
This study was conducted to evaluate the effect of an intelligent seat system, a microprocessorbased interactive seat that automatically adjusts itself to fit a seated individual by making pressure-sensitive adjustments on its own. Fist, a standard American automobile seat (‘baseline’ seat) was assessed for comfort. Subjective ratings of comfort, pressure distribution and seated anthropometric measurements were recorded for 20 test subjects. These measurements were recorded while the subjects maintained a simulated driving position in a seat buck. The comfort scale was based on a rating of 1 to 10, with 1 corresponding to ‘very poor’ and 10 corresponding to ‘very good.’ Based on a nonlinear, multiple regression model that had been previously developed, the comfort rating of the seat was predicted based on the subjective ratings and the recorded values of 450 pressure measurements from 20 subjects. The predicted comfort value was 7.46 for the baseline seat. Following the baseline assessment, the intelligent seat system was installed into the standard American automobile seat. The objective and subjective assessments were then repeated for 17 subjects and the new predicted comfort rating was 8.06. A t-test performed on the subjective and objective measures indicated that this was a significant improvement in seat comfort. Overall, subjects felt the self-adjusting seat was more customized and more comfortable, providing a better fit. Keywords: seat comfort, pressure distribution,
biomechanics
Prolonged sitting during daily driving activities can develop stress in muscles of the back, buttocks, and legs. Well-designed seating support can reduce this stress. When the body is not well supported, several muscle groups act together to restore stability, contributing to static loading. As a result, the occupant feels discomfort and fatigued. A well-designed seat should be able to accommodate all sizes and shapes of users and should provide adequate support (Gross et al, 1992). In this paper, we relate seat comfort to a biomechanical variable - pressure distribution between the seat surface and its occupant - and then describe enhancement of the comfort of the occupant by automatically adjusting the amount of support provided by the seat. Biomechanical research conducted on over 1100 seat-subject evaluations, comparing load distributions with subjective ratings, showed that comfort ratings can be predicted from patterns of weight distribution (Gross et al, 1994). This information allowed construction of multiple regression equations that predict comfort levels from seat pressure variables, leading to the development of the Intelligent Seat System (Figure I), a microprocessor-based interactive device that senses the pressures at the body-seat interface using sensors placed under the upholstery. Based on the occupant’s pressure distribution, the
AIR CEL
PRESSURE SENSING DEVICE
IIMICRO-+I I’
I-T-T
CUTOFF t!
II
PROCESSOR CONTROLLED II I
IIswTTcHMc IIL
Figure 1 Intelligent seat system schematic. High-pressure system, pressure measurement sensors, automated power inflate-power deflate on all bladders to optimize comfort, microprocessor-controlled switching. US Patent Number (1) 5,060,174 (2) 5,170,364 (3) 5,283,735
109
110
D. Ng et al
system decides which adjustments to make and automatically makes those adjustments by inflating embedded air bladders to optimize comfort. As a result, support and pressure distribution are improved. Over time, if the user changes posture or position, the Intelligent Seat System re-adjusts itself accordingly. A good distribution minimizes load concentrations that can restrict blood flow and affect the nerves, causing discomfort or pain. For example, pressure should not be concentrated primarily at the ischial tuberosities, but rather be distributed across the buttocks and thigh areas for better support. Sanders and McCormick (1987) suggest that the weight should be distributed rather evenly throughout the buttocks area, but minimized under the thighs. The objective of the study reported here was to evaluate the effect of the Intelligent Seat System after it was installed into a standard American automobile seat (called the ‘baseline’ seat). A comfort assessment was performed on the baseline seat without and then with the Intelligent System. Comparisons were made to see if there was a significant difference in seat comfort.
Table 1 Importance of seat features: percentage giving response (ranking from most important seat features to least important seat features) Ratings” Features
1
2
1. 2. 3. 4. 5.
65 70 35 70 60
3.5 30 60 30 20
Seatback recline Lumbar support Height of seatpan (from Seatpan tilt Head rest position
floor)
3 0 0 ;;20
“Ratings are based on the following descriptions: 1 - It is very important to me that this adjustability feature is provided 2 - This adjustability feature is not important to me, but I feel it should still be provided 3 - I do not care for this adjustability feature
Table 2 Importance of seat attributes: percentage giving response (ranking from most important attributes to least important seat attributes) Ratings’
Methods Subjects
The subject group consisted of 20 healthy individuals (10 male and 10 female), who were college students, skilled workers, and professionals. Subjects were independently recruited and hired. The average age of the subjects was 28.7 years, the average weight was 71.5 kg and the average height was 179.9 cm. The age of the subjects ranged from 19 to 51 years, with a mean of 28.7 years. The subjects’ driving experience level ranged from 3 to 35 years, with a mean of 12.1 years. A questionnaire was administered to the subjects to determine the important features of a car seat. Tables I and 2 summarize the subjects’ perception of the importance of seat adjustability features and seat attributes. As shown, 70% of the subjects felt that lumbar support and seatpan tilt are very important, while only 35% felt the height of the seatpan is very important. Also reported as very important were the seatback recline and headrest position. Subjects indicated that their perception of seat comfort was influenced most by thigh support (75%), thoracic support (70%), lumbar support (65%) and presence of armrests (65%). The seat was mounted on a seat buck, which simulated the driver’s side of a car (Figure 2). The seat buck was equipped with a steering wheel, a ‘dead’ pedal mounted at an angle of 40” for the left foot, and a simulated accelerator pedal (mounted at 60”) for the right foot. The baseline seat was mounted so that the front mounting bolt was 22.5 in (572 mm) from the heel point (accelerator pedal). The seat buck also had a clearance of 11.5 in (292 mm), which represented the height of the floorboard above the ground.
Features
1
2
I. Lumbar support 2. Seatback firmness 3. Texture and material of the upholstery 4. Seat cushion firmness 5 Thoracic support h: Presence of armrests 7. Seatback size 8. Buttocks support 9. Seatback lateral support 10. Thigh support 11. Physical appearance of the seat 12. Seat cushion size 13. Seat cushion lateral support 14. Head/neck support 15. Colour of the seat
65 50
35 45
0
2.5 60 70 65 50 60 15 75 50 IO 15 55 30
40 30 25 25 30 30 40 20 40 35 40 45 SO
35 10
3
IO
20 IO 10 JO
55 45 20
“Ratings are based on the following descriptions: 1 - This attribute of the seat greatly affects my perception of overall seat comfort 2 - This attribute of the seat somewhat influences my perception of overall seat comfort 3 - My perception of overall seat comfort is not at all affected by this attribute
Seat pressure measurement
Seat pressure measurements were obtained using thin flexible pressure mats (Figure 3) made up of forcesensing resistors. These pressure mats were configured over the seatpan and seatback and did not affect seat
Figure
2
Seat mounted on a seat buck simulating the driver’s
side of a car
Evaluation of an intelligent seat system
geometry. Two pressure mat configurations were used, one for the seatpan and one for the seatback. The seatpan mat was designed to cover an area of approximately 16 X 16 in (407 X 407 mm), which included 225 pressure points arranged into a 15 X 15 matrix. The seatback mat measured 14 X 21 in (356 X 533 mm) and also consisted of 225 pressure sensors arranged into a 15 X 15 matrix. At zero or very low pressures, the sensors performed as an open circuit. After the pressure had reached a low threshold, increasing force rapidly reduced electrical resistance. Data were collected from the seatpan and seatback pressure mats simultaneously with the subject seated at his or her preferred setting. During data collection, the subjects assumed a driving posture with both hands gripping the steering wheel, the right foot on the accelerator pedal and the left foot on the ‘dead’ pedal. Seat pressure data for the seatpan and seatback for each subject were obtained for the baseline and intelligent seat at a frequency of 45 Hz for 15 s. Subjects did not move throughout the entire interval. The averages of the seat pressure data for the seatpan and seatback were divided into 12 regions (Figure 4) for analysis. For each region, the mean and standard deviation of pressures, the sum of pressures, and the maximum and minimum pressure gradients were calculated.
111
‘.
-
Figure 3 Pressure mat consisting of 225 force-sensing resistors
Anthropometry
Sixeen seated anthropometry measures were recorded from each participant. All measurements were made using an anthropometer and a goniometer. These measurements were taken to ascertain the seated posture of the test subjects. Table 3 lists the ranges of acceptable body angles for a seated driver (Babbs, 1979). These figures are presented for comparison with the actual values reported in the results section.
1 9 1
Subjective measures of comfort
Data collection procedure
Each subject was briefed on the methodology and purpose of the experiment and then given a short questionnaire to determine level of driving experience and familiarity with different car seats. Subjects were also asked to rate generally the importance of certain seat adjustability features and attributes (as listed in Table 4). Subjects familiarized themselves with the seat and then adjusted it to their preferred seatings. After
6
7 6
5
Two questionnaires were designed to obtain subjective comfort ratings for various features of the seatback and seatpan. The first questionnaire required the subject to rate comfort on a scale of 1 to 10, with 1 representing ‘very poor’ and 10 representing ‘very good.’ The second questionnaire had a scale of +lO to -10 representing each extreme condition of the factor, with the midpoint 0 representing ‘just right.’ For example, for seat cushion firmness, -10 corresponds to ‘too firm’ and +lO corresponds to ‘too soft.’ These comfort ratings were based on a continuous interval with a normal distribution, as in previous studies we found it was consistent and mathematically more convenient to assume a normal distribution.
10
[
Seat Pan
L
1 Seat
Seat pan
Seat back
I 2 3 4 5 6
7 = R&t
= = = = = =
hght bolster Left bolster Right buttock Let? buttock hght thlgb Left thigh
Back
bolster
8 = Left bolster 9 = Right thora‘x
10=LeftthOraClC I I =Right lumbar I2 = Let? lumbar
Figure 4 12 regions of seat pan and seat back
Table 3 Recommended
rangesa for body angles for a seated driver
Angle
Range (degrees)
1. 2. 3. 4. 5.
la-45 m-120 9.5-120 95-135 90-110
Shoulder angle Elbow angle Hip angle Knee angle Ankle angle
“Babbs (1979)
D. Ng et al
112
achieving a feel for the seat, subjects were asked to rate the ‘comfort’ of each seat feature based on a scale of 1 to 10 (Table 5). Following the initial comfort rating, pressure data were acquired for the seat at the subject’s preferred setting and subjects were then given the second questionnaire with a rating scale of -10 to +lO for the features in Table 6. Seated anthropometric dimensions were then measured for each subject at the preferred settings. Adjustment ranges, physical measurements and characteristics of the baseline seat were also recorded (Tuble 4).
After the evaluation of the baseline seat, the Intelligent Seat System was installed. Air bladders were placed in strategic areas of the seatpan and seatback. A microprocessor-controlled switching device operated a reversible pump to inflate/deflate the bladders. After
Table 5 Mean ratings of comfort for the baseline and intelligent seat (with standard deviations). Scale: 1 = very poor to 10 = very good
Comfort
responses
1. Seatpan cushion
Table 4 Baseline seat characteristics
Light grey Leather surface Yes Yes 12 Electric Electric Electric Manual None Yes Manual
Colour Trim/texture Bolsters on pan Bolsters on back Number of adjustments Fore and aft control Back recline control Seat height control Lumbar adjustment Thigh bolsters control Adjustable seatpan Headrest Range of adjustment/dimensions Fore-aft (cm) (from gas pedal to front of seatpan) Min Max Seatback recline angle (degrees) (from horizontal) Min Max Seatpan height (cm) Min Max Height of lumbar curvature apex (cm) (above seatpan) Thickness of lumbar curvature (max) (cm) (at apex) Seatpan tilt (degrees) (from horizontal) Seatback with (lumbar area) (cm) (inside bolsters) Seatpan width (front) (cm) inside bolsters)
2 Seatpan width 3 Seatpan depth 4. Seatpan angle 5 Thigh support 6. Buttocks support 7. Overall seatpan support 8. Backrest cushion 9. Backrest height
25.4 48.3
10. Backrest width 11. Backrest recline
104” 150”
12. Lumbar support 13. Centre height/apex of lumbar support
27.9 33.0
14. Upper back support 19.1 15. Overall backrest support 3.91 16. Ease of accessing controls 10” 17. Ease operating controls 53.6 18. Overall comfort 29.9
Baseline (N = 20)
Intelligent (N = 17)
7.8 (0.98) 7.8 (1.28) 7.4 (1.20) 8.0 (1.46) 6.8 (1.32) 7.6 (1.06) 7.8 (1.02) 8.4 (1.08) 7.6 (1.48) 8.0 (1.20) 8.0 (1.34) 8.2 (1.24)
9.0 &92)
(1.48) 8.6 (0.86) 8.8 (1.16) 9.0 (0.90)
7.6 (1.36) 7.8 (1.42) 7.8 (1.18) 9.0 (1.34) 9.0 (1.46) 8.2 (1.04)
9.0 (0.86) 8.4 (1.70) 9.0 (1.06) 9.4 (0.80) 9.4 (0.88) 9.2 (0.76)
(1.42) 8.8 (0.88) 8.8 (0.84) 9.0 (0.94) 9.0 (0.90) 9.0 (0.94) 8.8 ;;08)
Table 6 Percentages of comfort ratings for the baseline seat and intelligent seat Feature/attribute
Baseline scat
Intelligent seat
1. Seatpan cushion firmness 2. Seatback cushion firmness
Too firm 20 20
Just right 55 55
Too soft 25 25
Too firm 11.8 29.4
Just right 76.5 70.6
Too soft 11.8 0
3. Seatpan height
Too high 10
Just right 70
Too low 20
Too high 5.9
Just right 88.2
Too low 5.9
4. Seatback size
Too large 0
Just right 75
Too small 25
Too large 5.9
Just right 76.5
Too small 17.7
5. Seatpan tilt
Too forward 10
Just right 85
Too backward 5
Too forward 0
Just right 94.1
Too backward 5.9
6. Seatback recline angle
Too reclined 10
Just right 85
Too upright 5
Too reclined 0
Just right 94.1
Too upright 5.9
Too much 10 20 10 0 1.5
Just right 40 60 35 30 40 55
Too little 50 20 55 70 45 45
Too much 17.7 5.9 29.1 5.9 0 0
Just right 82.3 88.2 58.8 35.3 88.2 82.4
Too little 0 5.9 11.8 58.8 11.8 17.7
7. 8. 9. 10. 11. 12.
Pressure on thigh area Lumbar support Upper back support Shoulder support Seatpan bolster support Seatback bolster support
0
Evaluation of an intelligent seat system
installation of the hardware and software was complete, 17 subjects from the first comfort study was recalled (3 of the 20 were unavailable) for a follow-up comfort assessment of the baseline seat with the intelligent technology incorporated into it. Seat geometry, cushion and configuration were unaffected. Subjects adjusted the seat to their preferred settings and then pressed a button to start the intelligent system. The system sensed the subject’s pressure distribution at the body-seat interface. Based on the pressure distribution, the seat then self-adjusted by inflating bladders in various areas (eg lumbar, side bolster) of the seat to provide more support, optimizing user comfort. After the seat had self-adjusted (about 35 s), ten samples of pressure data were collected and two subjective comfort questionnaires were administered.
113
on the intelligent seat. Figures 8 and 10 (seatpan pressure distribution) illustrate that the pressure was well distributed across the buttocks (55.6%) and thighs (37.4%). Figures 6 and 12 (seat back pressure distribution) display pressure as being most concentrated in the three regions: (lower, middle and upper lumbar) of the lumber area (76.5%). The rest of the pressure was distributed throughout the thoracic region and side bolsters,
Thorecic 16.24%
LUInbar bolsters 7.05%
Thoracic bolsters 2.30%
Results Pressure distribution
The load distributions for the seatback and seatpan regions are summarized in Figures 5-8. These values are expressed as a ratio of the pressure on an individual region to the total pressure on the entire seatpan or seatback and are used to determine whether proper support is provided to the occupant. Pressure plots were obtained by averaging the pressure distributions of all the subjects per seat and calculating the pressure gradient at each of the different points, Plots of the seatpan and seatback regions were made separately with contour intervals of 5% (data normalized to maximum values for the seat-subject combination, which was then assigned a value of 100%). Figures 9 and 10 show the average pressure gradient distribution for the seat pan and seat back for 20 subjects on the baseline seat. It can be seen from Figures 7 and 9 (seatpan pressure distribution) that pressure was not uniform across the buttocks and thighs. Most of the load was concentrated in the buttock region (73.9%), while 18.5% was found in the thigh region. Figures 5 and IO (seatback pressure distribution) show pressure to be most concentrated at the lumbar region (55.9%) and thoracic region (43.2%). Little pressure was found in the thoracic bolsters (0.2%) and lumbar bolsters (0.7%). Figures I1 and 22 show the average pressure gradient distribution for the seatpan and seatback for 17 subjects
Lumber 76.48%
Figure 6 Intelligent seat back pressure distribution
Pan Bolsters
Thighs 184%
Buttocks 73.64%
Figure
7
Baseline seat pan pressure distribution
Pan Bolsters Thoracic 4X22%-
bo~e~~~,%
Lumbar bolsters 0.68% Lumber 56.86%
Figure 5 Baseline seat back pressure distribution
Buttocks 56.66% Figure
8
Intelligent
seat pan pressure distribution
D. Ng et al
114 Subjective responses
The means of the subjective comfort responses before and after installation of the Intelligent System are listed in Table 5. The baseline seat had high ratings for ease of accessing and operating controls (9.0), backrest cushion (8.4), lumbar support (8.2), seatpan angle and backrest recline (8.0), and relatively low rating for thigh support (6.8). After installation of the Intelligent System, subjects’ perception of comfort increased substantially; for example, thigh support increased from 6.8 to 9.0, overall backrest support increased from 7.8 to 9.0 and lumbar support increased from 8.2 to 9.0. Table 6 shows rating of seat features and attributes from the second subjective comfort questionnaire for the baseline seat and intelligent seat. The percentage of subjects who rated ‘just right’ for pressure on the thigh
SEAT
PAN
PRESSURE
area increased from 40% of 20 subjects (baseline assessment) to 82.4% of 17 subjects (intelligent seat assessment). An increase in the ‘just right’ rating for the intelligent seat was also shown for the following features/attributes: lumbar support (60-88.2%), upper back support (40-58.8%), seatpan bolster support (4088.2%), and seatback bolster support (55-82.45%). Seatpan height, seatpan tilt and seatback recline angle were each rated as relatively high, possibly owing to subjects adjusting the seat to their preferred settings. Predicted comfort
A comfort model was used to predict seat comfort based on pressure distribution (Gross, 1992; Gross et al, 1994). The model was a multiple regression function that expressed the relationship between the
DISTRIBUTION 36
SEAT PAN PRESSURE DISTRIBUTION 0 2 4 6 7 9 11131517192022242628303133353739414344
33
g
33
30
30
27
27
24
24
$21
21
ln 18
18
x
15
w '5 0 m 12 z .P cc
orl’li J.l/r-J ’ ’ ’ ’ ’ ’ ” ’ I’IAiYi o 2 4 6 7 9 11131517192022242628303133353739414344° Front
of
Seat
SEAT 0
BACK
PRESSURE
6
6
3
3
0 0
- 27
27 i Y
- 24
:: 24 m
29 .g 6
10
Y
Seat
Pan
2
BACK
PRESSURE
DISTRIBUTION
4
6
7
9 11131517192022242628303133353739414344
4
6
7
9 1113151719202224262830313335
27
- 21
5
- 18
Ln"18
- 15
‘i; 15 Q,
21
- 12
w
-9
z9 .E"
-6
L1:6
12
Bottom
Back
Average pressure gradient distribution of all subjects for the baseline seat back
of
:: 24 m
9 11131517192022242628303133353739414344*
Seot
9 11131517192022242628303133353739414344°
SEAT
- 30
of
7
11 Average pressure gradient distribution of all subjects for the intelligent seat pan
0
133
Bottom
6
DISTRIBUTION
30
7
4
Figure
2 4 6 7 9 11131517192022242628303133353739414344 IIIIII//IIII
6
2
Front
33
Figure
g
”
9 Average pressure gradient distribution of all subjects for the baseline seat pan
36
12 9
Pa7
Figure
36
Figure
12
of
Seat
Bock
Average pressure gradient distribution of all subjects for the intelligent seat back
115
Evaluation of an intelligent seat system
dependent variable (comfort) and 17 independent variables (pressure). The final model was derived as a result of establishing the closest or best-fit relationship by using a combination of independent variables and eliminating those that contributed least to the comfort model. The pressure variables were obtained from the baseline seat and the intelligent seat. Pressure variables from previous car seat assessments were also included in the model to strengthen the correlation of comfort and patterns of pressure distribution. Based on a scale from 1 to 10, with 1 being ‘very poor’ and 10 being ‘very good’, and based on the values of pressure variables from the other seats, the predicted comfort value was obtained. The perceived comfort value reported here is an average of the 18 comfort responses provided by the test subjects from all 18 questions administered in the first questionnaire (Table 5). The predicted and perceived comfort values are presented in Figure 13; the comfort values for the intelligent seat (predicted = 8.06, perceived = 8.9) were higher than for the baseline seat (predicted =7.46, perceived = 7.9).
Table 7 Mean seated anthropometric measures (with standard deviations) for the baseline seat Measurements’
Mean
Sitting height
103.6
2. Elbow height
(4.6) 59.2
1.
3. Left knee height (2.9) 41.9
4. Right knee height 5. Left buttock-popliteal 6. Right buttock-popliteal
length length
(9.6) 86.7
8. Right ankle angle
(6.0) 134.9
9. Left knee angle knee angle
%.? it&;)
11. Left hip angle (6.1) 99.9
12. Right hip angle Subject anthropometry
Seated anthropometric data on all subjects are summarized in Table 7. These measurements were taken after the seat was adjusted to the preferred settings for each subject. The mean anthropometric values were compared with the recommended range of body angles listed in Table 3 (Babbs, 1979). Shoulder and elbow angles were found to be above the recommended ranges, possibly owing to seated postures in which subjects stretched their arms out more to reach the steering wheel. Hip, knee and ankle angles were found to fall approximately within the recommended ranges.
(3.1) 46.5 (2.3) 105.8
7. Left ankle angle
10. Right
(4.8) 48.8
13. Left shoulder 14. Right shoulder
(6.7) 49.0 (13.5) 52.1
angle angle
15. Left elbow angle
%2
16. Right elbow angle
(24.6) “All linear measurements measurements in degrees
in centimetres
and all angular
Conclusions Statistical analysis
A t-test was performed to determine whether a significant difference in comfort based on subjective results (Table 5) and objective results (predicted values) existed between the baseline seat and the intelligent seat. With a null hypothesis of no difference in comfort between the two seats, the results showed a significant difference in comfort (p = 0.001); that is, subjects felt significantly more comfortable after the intelligent system was installed.
12 10 8 6
4 2 0
Piedicted comfort
Perceived comfort
Figure 13 The predicted and perceived baseline and seat and intelligence seat
comfort
for the
The baseline seat had good ratings for ease of accessing and operating its controls, seatback cushion, lumbar support, seatpan angle and seatback recline. It was not well rated for thigh support and shoulder support. Most subjects liked the seat and lumbar adjustability features, and they found the lumbar support to be adequate. However, for most subjects (taller ones in particular) there was not much contact under the thighs, and they felt that more thigh support should be provided as well as a deeper seatpan depth. Subjects felt that the seatback and pan bolsters were inadequate, and that the bolsters were hardly noticeable. Even though the seatpan and back cushions had good ratings, most subjects felt that the cushions should be firmer. They also felt more lateral support should be provided in the seatpan and upper back. Pressure distribution data showed an uneven distribution across the thighs and buttocks, with most of the load concentrated in the buttock region. There was good pressure distribution in the lumbar and thoracic regions. Consistent with the subjective responses, loads on the thoracic and lumbar bolster regions were low, which indicates that more support should be provided in these areas. The comfort ratings for the intelligent seat were higher in almost all cases. Thigh support, upper back support, lumbar support, side and pan bolsters received a much better response compared to the baseline
116
D. Ng et al
assessment. Subjects rated the intelligent seat as more customized and accommodating, and providing a better fit. They particularly liked the way the seat adjusted automatically. Pressure distribution data showed better distribution across the buttocks and thighs compared with the baseline seat. In the seat back, pressure was mostly found in the lumbar area. This could be attributed to the fact that the three lumbar regions (lower, middle and upper) have provided greater support, relieving most of the pressure in the upper back. Common methods to distribute pressure in seat design are the use of contouring and seat cushion firmness. These produce a fixed shape, which does not provide easy adjustment to occupants of different sizes and weights, and as a result occupants may feel discomfort or fatigue during prolonged periods of sitting. An intelligent seat can avoid this by adjusting automatically to occupants of different sizes and weights, redistributing pressure and providing more support, thus optimizing user comfort and safety. This study concluded that subjects felt significantly more comfortable in the baseline seat after the Intelligent System was installed.
Acknowledgements The authors wish to thank the staff of McCord Winn Textron for their participation and support with the Intelligent Seat System, and acknowledge the excellent microelectronics design contributions for the Intelligent Seat System by Mr Jeff Finkelstein, of Microprocessor Designs, Shelburne, Vermont.
References Babbs, F.W. 1979 ‘A design layout method for relating seating to the occupant and vehicle’ Ergonomics 22, 227-234 Gross, C.M. 1992 ‘A biomechanical approach to ergonomic product design’ in Maynards Industrial Handbook 4th edn, MacGraw-Hill, New York, pp 8.45-8.61 Gross, C.M., Goonetilfeke, R.S. Menon, K.K., Banaag, J.C.N. and Nair, C.M. 1992 ‘Biomechanical assessment and prediction of seat comfort’ Automot Technol Int pp 329-334 Gross, C.M., Goonetilleke, R.S., Menon, K.K., Banaag, J.C.N. and Nab, C.M. 1994 ‘The biomechanical assessment and prediction of seat comfort’ in Hard Facts and Soft Machines: The Ergonomics of Seating Taylor & Francis, London, 1994 Sanders, M.S. and McCormick, E.J. 1987 Human Engineering Design McGraw-Hill, New York
Factors
in