Further to ideas expressed for determining travel patterns in this thread :
The difficulties are obvious as a reliable survey needs to be based on a large enough sample which will, in any case, be subject to a certain degree of error, besides high costs for such polls by private firms.
Instead, if BMTC /BMRC /BMLTA-DULT were provided periodically with inputs from various employers of a certain minimum size - based on the number of employees - say a minimum of 50 employees, it might yield better inputs at minimal cost.
An employer or firm employing over 50 persons need only fill out a sheet with some details of their employees such as name /designation, point of origin, destination, time of travel (to place of work & back), present mode of travel -- such data would be a lot more reliable & of course voluminous - almost a census, but once the procedures have been established, it can be done routinely on a five-yearly basis.
The advantages of such an exercise would be enormous for planning & meeting travel demand - It would give real information as opposed to extrapolation of sample sizes, as with polls. It would also reveal exact needs, particularly say, for bus routing & also for planning mass transits.
The difficulty is of course co-ordination & overseeing the exercise. The list of registered firms operating in the city should be available from government sources. If it were made mandatory for employers to file such information on a standard table format (say on the internet), the information could be very useful.
Perhaps praja folks, who are IT driven can come up with a program that easily processes the data & comes up with travel demand needs on each bus route !