Here's a rather unique problem statement. Slightly tangential but probably worth experimenting and exploring. Swiggy's in the nature of delivering food. In India (especially) getting food during a journey can be a bit cumbersome and a little bit tedious. It comes with a series of problems; is it healthy, is it trustworthy etc? Could it be possible to trial out delivering food at central bus terminals (only at boarding destinations). From restaurants that are in the vicinity. 

Here's what you'd be solving; a) trust b) hygiene c) randomness d) quality. 

For restaurants - another avenue to cater to users
For Swiggy - tapping a % of new users. 

Perhaps this product doesn't require any form of innovation, but more so some form of growth hacking?