Pizza maker in Cal­i­for­nia em­ploys 20 soft­ware en­gi­neers

The Hamilton Spectator - - BUSINESS - TRACEY LIEN SAN FRAN­CISCO —

Not long af­ter the pizze­ria Zume opened for busi­ness last year, its kitchen staff no­ticed a prob­lem with some of its piz­zas: they had holes in them.

It wasn’t the fault of the work­ers, who rolled out in­tact dough bases. There wasn’t a kitchen mole prod­ding holes. It wasn’t even the recipe — a Zume pizza base can han­dle its fair share of top­pings. It was the ro­bots.

Josh Gold­berg, 38, is the chief tech­nol­ogy of­fi­cer of the Moun­tain View, Calif., pizza joint. Al­though most pizze­rias don’t have an engi­neer­ing staff, let alone a CTO, Zume prides it­self on its use of au­to­ma­tion to make op­er­a­tions more ef­fi­cient.

It es­ti­mates its kitchen can make 10 times more piz­zas than a pizze­ria with a com­pa­ra­ble staff. It has a ro­bot that squirts tomato sauce onto its pies.

It has a ro­bot that spreads the sauce, mim­ick­ing the move­ments of Zume’s head chef. There’s a ro­bot arm (sim­i­lar to those found in auto man­u­fac­tur­ing fa­cil­i­ties) that puts the pie in the oven. And, as of this month, there’s a ro­bot that presses the dough into a per­fect cir­cle.

So if the com­pany has a non-hu­man prob­lem, it’s Gold­berg’s prob­lem.

Ob­serv­ing op­er­a­tions in the com­pany’s lab-like kitchen, Gold­berg watched as the hu­man cooks spun glossy blobs of dough and placed them on the con­veyor belt. He watched as a cam­era hov­er­ing above snapped a photo of the dough so it could in­form the other ro­bots of the pizza’s size, shape and pre­cise lo­ca­tion. An­other cam­era de­tected the cen­tre of the pie and in­structed a noz­zle to squirt sauce, and a delta ro­bot — the kind used on as­sem­bly lines — used a spiral move­ment to spread it. Hu­mans topped the pizza with pep­per­oni, fresh basil and cheese.

And just as the pizza was about to be put into the oven, Gold­berg found the prob­lem.

There was a small gap be­tween two con­veyor belts. If both kept mov­ing, then the pizza would glide along with­out any prob­lems. But if the con­veyor belts stopped for as lit­tle as two sec­onds, part of the pizza would sink into the gap, cre­at­ing a tear when the ma­chines started up again.

Fix­ing the prob­lem was easy enough. Gold­berg just had to pro­gram the con­veyor belt to not stop when a pizza was pass­ing over the gap.

But there are al­ways new prob­lems.

Ro­bots need cal­i­brat­ing, code needs to be up­dated. When­ever there’s a change in the dough or sauce recipe, the ro­bots must be taught new ways to work with dif­fer­ent tex­tures and con­sis­ten­cies.

That’s why Zume has a team of 20 soft­ware en­gi­neers. And its 20per­son kitchen staff doesn’t just prep in­gre­di­ents; many have been trained to work with the ro­bots. Its en­tire culi­nary team uses project man­age­ment tools such as Jira and Kan­ban, which are typ­i­cally used by soft­ware en­gi­neers at tech com­pa­nies for man­ag­ing projects and fix­ing bugs. And other staff, such as de­liv­ery driv­ers and line cooks, are be­ing trained in data sci­ence.

“We have fewer peo­ple, but we pay them much higher rates with full ben­e­fits as op­posed to hav­ing a pro­lif­er­a­tion of lower-skilled work­ers,” said Ju­lia Collins, Zume’s co-founder and co-chief ex­ec­u­tive.

The goal isn’t end-to-end au­to­ma­tion, Collins said.

There are still things that hu­mans do bet­ter than ma­chines, such as prep­ping in­gre­di­ents, mak­ing sauce, de­vel­op­ing recipes and know­ing when some­thing isn’t right with a pie. But au­to­ma­tion and soft­ware en­ables Zume to re­duce costs, make more piz­zas, pre­dict what piz­zas peo­ple want be­fore they or­der them and, even­tu­ally, take on the big pizza chains.


Jose Lopez pre­pares pizza dough to be placed on a con­veyor belt and on to an au­to­mated sauce dis­penser at Zume Pizza in Moun­tain View, Calif.

Newspapers in English

Newspapers from Canada

© PressReader. All rights reserved.