This is a question with different answers depending on the type of startup, the type of solution they are offering, their funding trajectory, and the founders' strategy and plans for growth.
Generally speaking, most startups burn cash in their early years. Cash flow positive can show up at year 3 or 4 on most hockey stick graphs, although the vast majority of startups don't have the hard data from the market to back up this projection, so any 5-year projections, as well as the cash-positive date, should be taken with a large grain of salt.
As a general rule of thumb, hardware and physical infrastructure startups with significant capital expenses (such as non-recurring engineering expenses and tooling and equipment costs, and in some cases, sitting / building costs) tend to take longer to get to cash flow positive than certain types of consumer-oriented software businesses with a capital-efficient cost structure.
That said, an enterprise software company selling large deals to large corporations will have a much longer time to cash positive due to the long sales cycle (and the long lead time for new field sales personnel to become productive in the event field sales is required - which remains the mainstream go-to-market method for selling large scale deals, including SaaS deals.
Hardware is tough, and Ben Einstein, GP at BoltVC (a hardware-oriented seed-stage VC and accelerator) has a very good article in 2015 on the financial picture behind a consumer electronics startup that remains highly relevant in 2019.
When you are talking about industrial level hardware you are in a whole new realm. Product delays can be costly because you can easily run out of runway before your product ships, so in order to succeed you will need to raise more than you think you need. For instance, the Tesla X is late by two years - they can survive this as they are very well funded, but a delay for a less well funded industrial hardware startup might not.
Generally, startups are not judged by how quickly they become cash flow positive - often there are other important considerations like proof of traction or growth of their subscriber base that factor into their metrics for success.
That said, startups are well-advised to think about a viable, scalable and sustainable business model from the get-go, and iterate it with customers over time. "Free" is not a business model and will not generally lead to a successful outcome (despite outlier cases like WhatsApp, Twitter and Facebook).
Special thanks: Martin Trust Centre, MIT
Edit: Javier Rojas, 12/05/20