There are many indicators that one could look at, "inverted yield curve" is one, but given it's popularity if it actually flips, it might actually trigger the recession in the behaviour that follows, so you end up needing to predict the predictor. And many indicators can get into "lucky socks" cargo cult territory quickly because they happen to match up events/stats with recessionary time periods.
If you need a general introduction, Dalio provides a nice starter for thinking about it:
https://www.youtube.com/watch?v=PHe0bXAIuk0
A interesting proposed measure here based upon fundamentals IMHO to me is consumption investment ratio:
http://necsi.edu/research/economics/econuniversal
All I can really say is be very sceptical of anyone who says they know for sure - I'm going mostly off my reading of cycle behaviors with some hand wavy harder measures (but economics is a terrible field and even the rigorously defined measures in the end have poor real-world predictive accuracy.)