The science and mathematics behind line balancing are always shunned by practicing managers, saying that mathematical calculations do not work with humane sensibilities and their inherent variability. Line balancing is therefore always practiced as an art and exploited by practitioners, playing into operator’s mind in an effort to improve efficiency and productivity through motivation, group incentives, and taking care of their welfare needs. Dr. Prabir Jana, NIFT Delhi discusses a simple mathematical technique that may improve line balancing efficiency and labour productivity.
Abalanced line maximizes productivity through maximizing operator and machine utilization. But balancing of line is dependent on exercising a control over operator’s skill inventory (hence operator rating), absenteeism and allocation.
Firstly, one must understand the dynamics of operator’s skill inventory and how it can help line balancing. As a primary role, this database maintains the record of each operator, who can do what operation and at what rating. It is very important to keep this database updated as over, the time operator acquires skills for new operations as well as improve performance in existing operations. Rating is the assessment of a worker’s rate of working. What the observer is concerned with is therefore the speed at which the operator carries out the work, in relation to observers’ concept of normal speed. It is necessary to have a numerical scale of rating to make the assessment effective. On a 0-100 scale, 0 represents zero activity and 100 – the normal rate of working of a motivated qualified worker i.e. the standard rate. Rating is used as factor by which the observed time can be multiplied to give the basic time. In operator’s skill inventory database, the rating should be updated regularly for near accurate balancing. For example, if the SAM/SMV of bottom hemming operation is 1.2 minutes (hence target production at 100 per cent will be 50 pieces per hour), and a sewing worker at bottom hemming operation has a rating of 80 per cent, which means the operator actually takes 1.5 minutes for bottom hemming and his/her production will be only 40 pieces per hour.
Usually factories use only one standard method to balance a line whereas many different scenarios can be weighed to choose from. Let’s begin with a shirt making factory with hourly target of 40 pieces per hour, three operations are having following details:
|Operation||SAM||Machine||Target per hour||Operator performance Required|
|A. Collar attaching||1.5||SNLS||40 pieces||100%|
|B. Cuff attaching||1.2||SNLS||40 pieces||80%|
|C. Band hemming||1.0||SNLS||40 pieces||66%|
Taken from operator’s skill inventory, the operator’s performances in aforementioned three operations are noted as under:
|For Operation A||70%||105%||140%|
|For Operation B||80%||100%|
|For Operation C||120%||70%||100%||70%|
On a peaceful ‘no-crisis’ day, the supervisor will allocate total three operators with closest performance available in the skill inventory list. For example in Operation A, we need 100 per cent rated operator, and closest available is Rohit (105 per cent), similarly for Operation B, we need 80 per cent rated operator and we have Govind and finally Shyam for Operation C. As shown in below table, we will be able to produce 42 pieces in Operation A and Operation C, and 40 pieces in Operation B. So we will be able to meet the target of 40 complete collars per hour and operator productivity for the section is 13.33 (40 complete collars divided by 3 operators). In the scenario, the services of Shyam have not been utilized, although he has the best rating in Operation A. Here, the logic of allocation is to find the closest match between operator performance required and operator performance available. This type of allocation results intrinsic balance of line. Intrinsic means the line balances itself.
Intrinsic line balancing operator allocation for Scenario 1:
|Operation||Intrinsic Balance||Actual Rating||Actual Production (pieces/hour)|
|A. Collar attaching||Rohit||105%||42|
|B. Cuff attaching||Govind||80%||40|
|C. Band Hemming||Shyam||70%||42|
Let’s take another day in the same factory sewing the same style. But due to the festival season there is heavy absenteeism and Rohit and Govind both are absent. We have to achieve best production possible using only two operators. Now we have to use different balancing logic, where the target is to utilise the operators in operations they can do best. This approach results dynamic balance of line. We will use Shyam and Rehman and allocation of work will be like: Shyam will be working 37 minutes in Operation A and 23 minutes in Operation C and can sew 34 pieces of Operation A and 16 pieces of Operation C, respectively. Similarly, Rehman will also share his time between Operation B and Operation C for 42 minutes and 18 minutes, respectively. With two operators we can do 34 pieces per hour and even during absenteeism with smart balancing logic operator productivity rise to 17 (34 complete pieces divided by 2 operators).
Dynamic line balancing operator allocation for Scenario 2:
|Operation||SAM||Operator Allocation (minutes worked)||Rating||Production (pieces/hour)|
|A. Collar attaching||1.5||Shyam (37)||140%||34|
|B. Collar topstitch||1.2||Rehman (42)||100%||35|
|C. Collar runstitch||1.0||Shyam (23)||70%||16|
On day 3, Shyam, Rehman and Govind all three are present. Encouraged by the last experience of dynamic line balancing the supervisor tries the same logic to utilize the operators in operations they can do best. Shyam is given Operation A completely, Rehman Operation B completely and Govind shared his time between Operation B and Operation C (in the ratio of 10 minutes and 50 minutes, respectively, in every hour). With three operators we did total 56 complete pieces per hour and operator productivity rise to 18.66.
Dynamic line balancing operator allocation for Scenario 3:
|Operation||SAM||Operator Allocation (minutes worked)||Rating||Actual Production (pieces/hour)|
|A. Collar attaching||1.5||Shyam (60)||140%||56|
|B. Collar topstitch||1.2||Rehman (60)||100%||50|
|Govind (10)||80%||6 (actually 6.67)|
|C. Collar runstitch||1.0||Govind (50)||120%||60|
The interesting question arises is to why factories may not be very keen to practice this? Although this dynamic line balancing results in better operator productivity, it is comparatively difficult to maintain by supervisor. It has operational complexities like movement of operator is more, warming up loss is more. Mathematically it may be ‘OK’ to say Shyam should work for 37 minutes and 23 minutes, respectively in two operations, however in reality it is difficult to maintain such odd timings. There is a feeling that any allocation of less than 60 minutes may require additional acclimatizing time (resulting in warming up loss) for operators, which may negatively impact production.
Even if we factor in the warming up loss during every movement of operator from one operation to another, the dynamic balancing will clearly give better results. Only plus point for intrinsic balance is comparatively peaceful time for line manager as once allocation is done for the day, not much shuffling (re-allocation) is needed, except in exigencies like machine breakdown, medical time-out by operator, unexpected quality problems, etc. On the other hand in dynamic line balancing, manager will have a tough time continuously re-allocating operators as the cases of one operator doing more than one operation will increase substantially. On the positive side this will give better results during heavy absenteeism and operators also will earn more, as they are working on operators where they can do best. A quick comparison of intrinsic and dynamic line balancing is as under:
|Parameters||Intrinsic Balance||Dynamic Balance|
|No. of people required||More||Less|
|Movement of operators||Less||More|
|Heavy absenteeism||Not preferred||Preferred|
|Supervisor skill||Less headache||More headache|
|Operator utilization||Not so good||Better|
It is high time our managers appreciate the logic of line balancing, do some homework (instead of guess work) and harness the best from the workforce.