One thing I use when running Supercrew is the system of Villain Points developed by a member of RPG.net. These allow you to set a challenge, be it a gang of thugs, a super-villain, and earthquake or simply finding a place to park, at a particular level scaled to the number of players. A GM can decide on the difficulty of the challenge, and put together the opposition then and there. It's very easy to do this on the fly.
When playing solo it can be obvious when a challenge arrives. The difficulty comes in getting the 'automated' GM to set an appropriate level. This is the system I use.
Using RPGSolo The game is run by means of a series of Yes/No questions - basically the questions a player or players would ask a live GM. The answers can be as follows:
Yes, And ... - This means the answer is yes, and there's some bonus as well ('Yes, there is a door, and it's unlocked!')
Yes - Exactly what it says.
Yes, But ... - This is an affirmative answer, but there is a negative rider of some kind.
No, But ... - The opposite of the above; the answer is No, but it's qualified in a way that is slightly positive to the players.
No - No means No.
No, And ... - This is bad - the answer's No, and there's more. Usually not good.
The probability of getting a 'Yes' answer is given a likelihood, and this will weight the responses accordingly. The probabilities range as follows:
When determining the difficulty of a challenge I ask the question 'Is the Challenge Easy?'. Always. What varies is the likelihood that it will be easy. I will start of by assuming that it's Likely or Somewhat Likely that the first challenge will be easy, because you don't want things to get too tricky straight off. The system takes over from there. This is how I read the results:
Yes, And ... - The challenge is Easy, and there is no change in probability.
Yes - The challenge is Easy, but the likelihood that the next one will be goes down one level.
Yes, But ... - The challenge is Easy, but the likelihood that the next one will be goes down two levels.
No, But ... - The challenge is Standard, but the likelihood of the next one being Easy goes up one level.
No - The challenge is Standard. There is no change in the likelihood that the next one will be Easy.
No, And ... - The challenge is Hard, but the likelihood of the next one being Easy goes up one level.
It looks complex, but isn't. Here's an example:
Our Hero encounters a challenge. I ask 'Is it Easy?' and decide that it's Likely. I get 'Yes, But' as the answer. So the challenge is, indeed, Easy. Buy the likelihood that the next one will be Easy goes from Likely to 50/50.
The next time our Hero encounters a challenge I ask 'Is it Easy?'. It's now 50/50, and I get 'No'. This means that the challenge is Standard, and the likelihood that the next one will be Easy stays at 50/50.
A third challenge arises. Again I ask 'Is it Easy?', and this time I get 'No, But'. So again it's a Standard challenge, but the likelihood of the next one being Easy improves from 50/50 to Somewhat Likely.
So basically the more Easy challenges you get, the more chance there is of getting one that's Standard or Hard.
Needles to say I haven't fully tested this yet, but it's worked out OK so far.