Yesterday marked the expiry of US Patent 5597589, "Apparatus for producing parts by selective sintering." This is one of the core patents in the 3D printing world -- the patent that allows 3D printer companies to charge more for fine nylon powder than Michelin-starred restaurants charge for filet mignon. The high cost of consumables in 3D printing has been a major barrier to innovation in the field -- selective laser sintering produces a fine finish that the patent-free fused deposition modeling technique used in Reprap-style printers can't match -- and now the brakes are coming off.
However, there are still lots of patents (including some genuinely terrible ones) in the 3D printing world, so the expiry of 5597589 doesn't necessarily mean that we'll see a flood of cheap printers and cheaper feedstock -- given the murkiness of the overlapping patent claims and the expense of litigating each one of them, radical new entrants into the field are still facing a lot of risk that has nothing to do with making great products at a fair price.
In a good piece on 3D Print, Eddie Krassenstein speculates about the scary supplementary laser-sintering patents lurking in the wings, pointing out that Stratasys (the major competitor of 3D Systems, who owned 5597589) didn't design their entry-lever printers to use SLS, even though they knew that the patent would be expiring in early 2014. Krassenstein suggests that this means that Stratasys knows about some other gnarly and deadly patent that would torpedo them if they went SLS.
But I'm a lot less convinced than Krassenstein is about the potential of a competitor taking the risky step of making a SLS printer that sticks to the claims in 5597589. Virtually every technical idea is covered by a stupid, overbroad patent, and yet people start businesses every day that open them to legal liability from a troll or an entrenched incumbent. If the potential for a patent suit was, in itself, a sufficient deterrent to raising capital and starting a business, we wouldn't see any startups. And a company that sticks to the claims in 5597589 has a powerful weapon in any patent suit: the USPTO granted 5597589 20 years ago, and so if they granted overlapping patents since, they were manifestly in error, a matter that is relatively (in patent terms, anyway) easy to prove.
Read the rest
Read the rest