2019-02-11

Planning & Strategy.

• Calling their customers "entrepreneurs," Patreon's CEO Jack Conte says that the company's business model (a 5% fee of their customers' sales) is insufficient and that "the reality is Patreon needs to build new businesses and new services and new revenue lines in order to build a sustainable business." Patreon has raised $105.9M since 2013, compared with Kickstarter's $10M raised since 2009 and Indiegogo's $56.5M raised since 2008; those discrepancies would presumably bring a different set of investor pressures. As a Patreon user (on both sides of the marketplace) myself, I'm not particularly sympathetic to Comte's arguments: Their creator tools are simply not suited for anything resembling "entrepreneurship," and in my interactions with Patreon employees I see no indication that they take that shortcoming seriously.
• Tesla will acquire Maxwell Technologies, which “has struggled for decades to find a way into the automotive market” and makes lithium-ion capacitors.
• Craig Cannon on the podcast ecosystem (in particular the lack of discoverability) in the context of Spotify's recent acquisitions. See also my 2017 interview with Craig, which touched on similar subjects.

Making & Manufacturing.

• A pretty marvelous old photo of a centrifugal compressor wheel being machined on a Cincinnati dial-type universal mill; note the tracers on the left side and scroll through the comments to get a sense of how revered these machines are. Parts like this one remain a favorite for anyone looking to show off their manufacturing skills, including the big metal printing companies today. I also recommend browsing through the lathes.co.uk page for Cincinnati - the photos (as always) are really beautiful.
• A robotic cookie machine that allows the user to tweak the recipe to find optimal ingredient ratios.
• An update on Etna Tool & Die, the famed "last tool & die shop in NYC."

Maintenance, Repair & Operations.

• Tesla's repair parts operation is, as you might imagine, woefully underdeveloped.
• A good profile of NYC's porta-potty industry, which has begun to be absorbed into nationwide chains. "Portable toilets are a lagging economic indicator. In a recession, housing starts drop, but projects under construction are typically finished. Porta-potties stay on site until the bitter end. But toilets can also be a canary in the coal mine. Fewer events means fewer opportunities; no new construction means no new toilet orders."

Distribution & Logistics.

• Instacart had been shorting their delivery contractors' pay by including tips in base pay.
• Apparently Japan allocates funding for their sewers partly by charging residents/companies for the waste they flush and partly by levying taxes to cover the cost of storm water. "Therefore, municipalities with more combined sewers receive more money from general tax than those with only sanitary sewers." This strikes me as a perverse incentive if the goal is to encourage sanitary sewers, which have big benefits in public & environmental health. If anyone has more details on how the Japanese system works, please send.
• TransitCenter is launching a Women Changing Transit mentorship program.

Inspection, Testing & Analysis.

• Thanks to everyone for responding to my question about mean time between failure of mechanical & computer parts last week. The consensus was that due to a) the relatively short history of computers, b) the fact that we live (and die) in a mostly mechanical context, and c) the high rate of change in computer equipment (and therefore the tendency to scrap parts that are out of date, regardless of whether or not they're actually broken), our understanding of how mechanical stuff fails is *probably* more well developed. Two good links on the computer side, though: One paper on errors in DRAM and one on hard disk failures, which found relatively low correlations between operating temperatures and failure rates. See also this good teardown of LED lightbulbs, with explanations on both the engineering and business reasons why they typically fail well before the 100,000 hours we were originally promised they'd last.
  
   I'll add a hypothesis here: Knowledge about the failure modes & lifetimes of mechanical components is more likely to exist in national and international standards bodies (ASTM, ISO, ASME, etc) than similar knowledge about computer components (which is often owned by individual companies). The corollary, of course, is that systems for improving reliability & decreasing failures in mechanical components are less likely to be open source & free for anyone to access.
  
   As someone who works in both the physical and digital worlds, I'd love to develop these ideas; get in touch if you have thoughts.
• John Dimatos on the history of crowdfunded micromobility projects.

Tangents.

• Email from an overconfident white dude. "If you're struggling with imposter syndrome, please remember that this guy isn't, and channel some of his chutzpah."
• Angela Ahrendts will leave her post as head of retail at Apple.
• Hagfish slime is *crazy.*
• I continue to work through the Elena Ferrante books, which are nothing if not emotionally engaging, on audio.

An interactive tool for exploring Manhattan's daily/weekly population migration patterns.