Hacker News Reader

Comment by nkozyra | Hacker News Reader

To me, food preparation is the big void in automation. A few companies are trying it here and there, but being able to prepare a decent-tasting meal (even if you have to do the cleaning yourself) would change some lives dramatically.

These aren't things that necessarily require advanced AI or sensors or anything; the basic elements of cooking are relatively simple. The real issue is the vast diversity of cooking/compilation methods. In other words, automating meals could be done, but automating all meals would require something monumental.

But consider simple baked meals, which comprise thousands of recipes.

There are people in the space, but I think a commercial grade Cookbot is a holy grail for home automation.

View on HN

Time use studies support what you're saying:

https://www.bls.gov/tus/charts/household.htm

The most time consuming activity in households is food preparation.

The other way that cookbots can be addressed is by preparing food in kitchens and then shipping it out at very low cost. Autonomous vehicles are ideal for that. The food can then either be prepared by people or robots in commercial kitchens.

Robots are not very good at manipulating things, especially things that are flexible or things with a variety of shapes. So one of the biggest problems that ingredients don't come in standardized containers. So either one has to sell every ingredient in a special cartridge or the human has to load ingredients into the machine. Now since some ingredients are perishable this makes loading the machine weekly and producing cartridges difficult. If a human has to manually load the machine then the human is still doing quite a bit of labor. Now making simple baked meals such as casserole are harder than they look for robots. So we can probably mix ingredients together, we might be able to slice vegetables, but getting an mixture of ingredients into a baking pan might be difficult. Even simpler is stir fry, and there are already robots that do this[0] In addition, it is also worth pointing out that devices that manipulate granular materials(flour, rice, salt, etc) are not very reliable[1]

[0]https://www.businessinsider.com/mit-students-invented-a-robo... [1]https://www.rand.org/pubs/reports/R3216.html

I don't understand why freeze dried ingredients aren't more common. I understand it takes considerable energy, but it seems like one of those problems of scale that you could throw money at, like batteries, and really change the market dynamics for considerable profit.

With freeze dried ingredients you could much more easily create standard containers to hold chopped vegetables and protein. It's a short step to then create a machine that can create a diverse array of sauces on demand; and many cooking styles are centered around sauces. Casseroles would work similarly.

There's a niche market for freeze dried foodstuffs that caters to the survivalist community, but it's expensive. I think most preppers buy traditionally preserved items--heat drying, salted, etc. Bringing down the cost of freeze drying such that you could enable new markets would require tremendous scale, but the money is there. We just need an Elon Musk-type person to put the cash to work.

I guess I should point out that freeze drying foodstuffs substantially preserves taste and, to a lesser extent, texture. Nothing else comes close. I have a cabinet full of freeze dried, chopped herbs, including red onions and garlic; they're simply incomparable to regular dried herbs and essentially obviate the need to buy (and plan ahead for) fresh herbs. Freeze drying preserves the volatile compounds that provide depth of flavor.

I LOVE cooking, there is no way to express how much I love it. But strangely this love also has the implication that I hate when I need to cook fast. Sometimes I hang out a lot in the gym or work and come home around 9-10pm. At that point cooking for an hour is not option as it's almost bed time, so I cook fast. I hate this, and I just want ready food (but I never eat out as a principle).

I searched into this, tried to find startups that can automate my food. I'd be willing to pay a lot of money if a robot made my food exactly the way I want it, and then I can cook long meals 3 to 5 times a week. But unfortunately, the only relevant products are liquid food. There are some startups, such as Soylent, who claim their product has complete nutritional value. So you just shake their powder in water and be done with dinner in 2 minutes. I personally thought this is a distraction from the problem, but I honestly don't know how could we automate food.

Take a look at Thermomix[1]. It's a food processor with a heated bowl and built-in scale, and the range of foods it can cook is surprisingly wide, given some help from the user. You can cook chicken dishes, lasagna/moussaka, steam vegetables and all sorts of things.

BTW, their sales model is quite unique. There probably are no retail stores near you. Instead, they come to your home and they prepare some food for you to show you what it can do, and then you have the option to buy it.

[1] https://thermomix.com/

Yeah. I enjoy cooking a lot too but that doesn't mean I want to do it every night of the week. In practice, I can plan things I can whip up quickly and there are some frozen foods (or leftovers) I can heat up. But it would be nice if I had options when I just want to eat something. (I have essentially no takeout or delivery options where I live.)

Why not batch cook and plan on sometimes reheating leftovers?

Because I think both the process of batch cooking and reheating are bad for the flavor.

some planning plus a timed "slow cooker" could address this directly

> Because I think both the process of batch cooking and reheating are bad for the flavor.

People who make a point of sticking to their preferences seldom take advice.

There’s a restaurant in Boston that has a robotic kitchen. It’s called Spyce [0], and I’ve seen them post in Ask HN who is hiring threads here. I’m not affiliated with them at all, but it seems like an interesting idea.

[0]: https://www.spyce.com

Watching the first half of their video, what they seem to have created isn't a "robotic kitchen" but the equivalent of a mechanized ladle. It looks like humans still do all of the ingredient preparation, mixing, garnishing and serving, and the "robots" are simply self-stirring self-washing pots. Which is still cool, don't get me wrong, but nowhere near the real thing.

I think you're imagining something with arms you bolt onto the counter & talk to.

But maybe it doesn't look like that. Maybe it's better for the robot to work 24/7 to pay off its cost, by serving many people. It can collaborate with other robots specialised in different tasks. And this we already have, it's called a factory. You can buy an enormous and ever-growing variety of partially prepared foods, where the robots have done much of the work for you.

How much of the work, well, it depends on your starting point. Ever shelled peas or plucked a chicken? Few people now do that at home, but it was once essential... our grandmothers would be astonished how little work is now needed.

Apartments could have a central robot kitchen and dumb waiters to send it up. Would be nice and easy to order simple stuff like pasta and pizza. Or also delivery of basic items such as milk. I'd love not having to carry heavy stuff home, or worry about if I have enough at home in the first place.

I think you could automate a wide variety of dishes by making a robot that dumps ingredients from hoppers into a food processor then dumps the chopped food into a skillet and stirs them. You'd manually peel/core/debone your ingredients, tell it what you put in each hopper, and let it figure out the chopping level, heat, and cooking order.

Do you need a robot for that, or an appliance? All these sous vide startups and the Instant Pot have come a long way towards automating their respective niches.

At this point, the only thing that you could really automate is stirring a pot. Everything else is either too complicated or too subjective (how rare a steak should be, etc.)

This is basically the “take home meal kit” that many different companies offer now. It’s too expensive. When you add in the labour, you might as well just go out for dinner.

Cooking the meal is the cheapest part, prepping is the most expensive.

> Cooking the meal is the cheapest part, prepping is the most expensive.

That's ... sort of my point.

That's what the Thermomix and its ilk are for. They're extremely popular in some countries.

Reply on news.ycombinator.com