4 ms·

I think the business case with the same idea and tech, but a lot better market cap, would be providing a restocking service to businesses who are using tens of

I think the business case with the same idea and tech, but a lot better market cap, would be providing a restocking service to businesses who are using tens of thousands of dollars of consumable shop supplies than the small audience of people who are willing to buy coffee from a subscription for $3/mo. To make the pain point that I'm familiar with on a daily basis crystal clear to you: I'm an engineer at a small, shop that makes industrial automation equipment. Each machine contains (in addition to bill-of-materials listed components, an array of stuff we consider "shop supplies". Some components I'd inventory include a lot of items from the McMaster Carr, Grainger, Bolt Bin, and Digikey catalogs. We have racks of pneumatic fittings, electrical conduit connectors, dowel pins, electrical connectors, extrusion hardware, labels, and all manner of other industrial Lego. There's a large area with hundreds of drawers holding alloy cap screws in socket, button, and flat-head in metric and SAE from #4-40/M3 to 1/2-18/M10 in quarter inch/5mm increments. Every tool and die shop, industrial facility maintenance department, auto mechanic, construction contractor, and so on has a similar set of drawers, shelves, and ULine bins that contains stuff like this. Each item only runs a nickel to a couple dollars each, so it's hardly worth adding to a spreadsheet, checking out of inventory, and cutting a purchase order for a replacement, but our little 20-man shop probably goes through $150,000 worth of these consumables each year in the process of building $10M worth of machines. Keep our fabricators supplied with the components they need, and stop our expensive engineers from spending $50 in time every week or two purchasing these components, and I'd be more than happy to pay for the value it would provide. Which is a lot more than the value that would be provided by keeping my home kitchen stocked with coffee. Also, from my point of view, the equipment to use to measure this stuff is not ESP8266 modules and a soldering iron, it's an off-the-shelf PLC, some DIN rail and screw terminals, and some load cells or maybe strain gauges if you really want to optimize costs. Get a volume of equipment sold before trying to roll your own full-stack hardware supply chain...but that's beside the point.

mooman219 8y ago
This would be killer for the workplace and tracking these incidental materials without spending people hours to do so. At home, if this is on a per-item basis, I could see you being able to set expiration timers too, and triggering a re-order then.
seizethecheese 8y ago
Thanks for laying out the use case! I’ve previously worked in restaurants and there’s an analogous problem there. I think this tech will be everywhere someday. I’mcurious about your insight on the tech. Could you flesh that out a bit more?
- LeifCarrotson 8y ago
  Sure. Essentially, the parent article was discussing how they broke out the soldering iron and learned about wireless integrated circuits. At a large enough scale, it makes great sense to do this engineering work: design a PCB with power supplies and analog circuits, figure out connections to peripheral sensors, put that PCB and sensors into an injection-molded piece of plastic, add fasteners and brackets to mount sensors, install connectors and terminate wires on the sensors and boards, write assembly and C code to make them all talk to each other ...taking on all the overhead. But for a lot of small runs of equipment, this doesn't make sense. Buy the strain gauge already in an enclosure with standardized interfaces and built-in amplifiers, get a small computer (PLC) that's already in an enclosure and has terminals to connect to those sensors, write code in a high-level language designed for talking to sensors, and ship it. It's like somebody wanted to build a website, so they started digging into network protocols. There's a place for that, but it should be approached long after you've fired up a Wordpress instance.
peterwwillis 8y ago
Came here to echo this, but for grocery stores ala Amazon's bumbled attempts to optimize Whole Foods stocking. Your use case and the parent's sound like a good use of JIT Kanban[1], waiting for inventory levels to drop below a threshold to replenish them. However, demand-forecasting[2] may also be useful in the parent's use case, and sometimes both systems are used in a product lifecycle. [1] https://en.wikipedia.org/wiki/Kanban https://en.wikipedia.org/wiki/Kanban [2] https://en.wikipedia.org/wiki/Demand_forecasting https://en.wikipedia.org/wiki/Demand_forecasting

mooman219 8y ago

This would be killer for the workplace and tracking these incidental materials without spending people hours to do so. At home, if this is on a per-item basis, I could see you being able to set expiration timers too, and triggering a re-order then.

seizethecheese 8y ago

Thanks for laying out the use case! I’ve previously worked in restaurants and there’s an analogous problem there. I think this tech will be everywhere someday. I’mcurious about your insight on the tech. Could you flesh that out a bit more?

LeifCarrotson 8y ago
Sure. Essentially, the parent article was discussing how they broke out the soldering iron and learned about wireless integrated circuits. At a large enough scale, it makes great sense to do this engineering work: design a PCB with power supplies and analog circuits, figure out connections to peripheral sensors, put that PCB and sensors into an injection-molded piece of plastic, add fasteners and brackets to mount sensors, install connectors and terminate wires on the sensors and boards, write assembly and C code to make them all talk to each other ...taking on all the overhead. But for a lot of small runs of equipment, this doesn't make sense. Buy the strain gauge already in an enclosure with standardized interfaces and built-in amplifiers, get a small computer (PLC) that's already in an enclosure and has terminals to connect to those sensors, write code in a high-level language designed for talking to sensors, and ship it. It's like somebody wanted to build a website, so they started digging into network protocols. There's a place for that, but it should be approached long after you've fired up a Wordpress instance.

peterwwillis 8y ago

Came here to echo this, but for grocery stores ala Amazon's bumbled attempts to optimize Whole Foods stocking. Your use case and the parent's sound like a good use of JIT Kanban[1], waiting for inventory levels to drop below a threshold to replenish them. However, demand-forecasting[2] may also be useful in the parent's use case, and sometimes both systems are used in a product lifecycle. [1] https://en.wikipedia.org/wiki/Kanban https://en.wikipedia.org/wiki/Kanban [2] https://en.wikipedia.org/wiki/Demand_forecasting https://en.wikipedia.org/wiki/Demand_forecasting