Singapore-based FinAccel raises $90M Series C for its credit lending app Kredivo, bringing its total equity fundraising to $140M — Singapore-headquartered FinAccel has secured $90 million in what is the largest funding round for a fintech startup in Southeast Asia as it looks to further grow …
Even to an experienced carpenter, it may not be obvious what the best way is to build a structure they’ve designed. A new digital tool, Carpentry Compiler, provides a way forward, converting the shapes of the structure to a step-by-step guide on how to produce them. It could help your next carpentry project get off the screen and into the shop.
“If you think of both design and fabrication as programs, you can use methods from programming languages to solve problems in carpentry, which is really cool,” said project lead Adriana Schulz from the University of Washington’s computer science department, in a news release.
It sounds a bit detached from the sawdust and sweat of hands-on woodworking, but they don’t say “measure twice, cut once” for nothing. Carpentry is a cerebral process more than a physical one, and smart, efficient solutions tend to replace ones that are merely well made.
What Carpentry Compiler does is codify the rules that govern design and carpentry, for example what materials are available, what tools can do, and so on, and use those to create a solution (in terms of cuts and joins) to a problem (how to turn boards into a treehouse).
Users design in a familiar 3D model interface, as many already do, creating the desired structure out of various shapes that they can modify, divide, pierce, attach, and so on. The program then takes those shapes and determines the best way to create them from your existing stock, with the tools you have — which you can select from a list.
Need to make the roof of your treehouse but only have 2x4s? It’ll provide a recipe with that restriction. Got some plywood sheets? It’ll use those, and the leftovers contribute to the base so there’s less waste. By evaluating lots and lots of variations on how this might be accomplished, the program arrives at what it believes are the best options, and presents multiple solutions.
“If you want to make a bookcase, it will give you multiple plans to make it,” said Schulz. “One might use less material. Another one might be more precise because it uses a more precise tool. And a third one is faster, but it uses more material. All these plans make the same bookcase, but they are not identical in terms of cost. These are examples of tradeoffs that a designer could explore.”
A 24-inch 2×4 gets cut at 16 inches at a 30-degree angle.
That’s really the same kind of thing that goes on inside a woodworker’s brain: I could use that fresh sheet to make this part, and it would be easy, or I could cut those shapes from either corner and it would leave room in the middle, but that’ll be kind of a pain… That sort of thing. It can also optimize for spatial elements, if for example you wanted to pack the parts in a box, or for cost if you wanted to shave a few bucks off the project.
Eventually the user is provided with a set of instructions specific to their set of tools. And the carpenters themselves act as the “processor,” executing operations, like “cut at this angle,” on real-world materials. In Carpenter Compiler, computer programs you!
The team presented their work at SIGGRAPH Asia last month. You can read more about the project (and learn how you can try it yourself) at its webpage.
Facebook created a chatbot called “Liam Bot” that teaches employees official company talking points for responding to friends and family critiquing the company — The “Liam Bot” teaches employees what to say if friends or family ask difficult questions about the company over the holidays.
In the United States, electricity generation accounts for nearly 30% of our carbon emissions. Some states have responded to that by setting aggressive renewable energy standards; others are hoping to see coal propped up even as its economics get worse. Complicating matters further is the fact that many regional grids are integrated, meaning power generated in one location may be exported and used in a different state entirely.
Tracking these electricity exports is critical for understanding how to lower our national carbon emissions. In addition, power from a dirty source like coal has health and environment impacts where it's produced, and the costs of these aren't always paid by the parties using the electricity. Unfortunately, getting reliable figures on how electricity is produced and where it's used is challenging, leaving some of the best estimates with a time resolution of only a month.
Now, three Stanford researchers—Jacques A. de Chalendar, John Taggart, and Sally M. Benson—have greatly improved on that standard, and they have managed to track power generation and use on an hourly basis. The researchers found that, of the 66 grid balancing areas within the United States, only three have carbon emissions equivalent to our national average, and they have found that imports and exports of electricity have both seasonal and daily changes. de Chalendar et al. have found that the net results can be substantial, with imported electricity increasing California's emissions/power by 20%.
Read 10 remaining paragraphs | Comments
Want to get professional feedback about your pitch deck, marketing assets, or immigration issues? Attending Disrupt Berlin?
We’ve brought in experts from around Europe (and Silicon Valley) to help you directly.
These events were hugely popular when we first did them at Disrupt in San Francisco the other month, so get your submissions in ASAP.
The Top 3 Immigration Mistakes Startups Make
11:25 – 11:35, Wednesday, Dec. 11
Sophie Alcorn (Alcorn Immigration Law)
Sophie Alcorn has helped hundreds of startups worldwide navigate the complexities of United States immigration law. She’ll be leading a talk on top mistakes, participating in a panel on global scaling, and providing 1:1 immigration advice sessions for any attendee who is interested.
You can now sign up for a time slot with her using our CrunchMatch offering. Attendees can request appointments through CrunchMatch in the TechCrunch events app, available upon registration.
Pitch Deck Teardown
16:45 -17:25, Wednesday, Dec. 11
Russ Heddleston (DocSend)
Karen Stafford (Intel)
Sitar Teli (Connect Ventures)
How can you make your deck bring in the funding? Two top European investors and a startup founder with a special data set about will be reviewing submitted decks live on the Extra Crunch stage.
Want to see what they think about yours? Just fill out this form.
Growth Marketing 2020
14:05 – 14:45, Thursday, Dec. 12
Asher King Abramson (Demand Curve)
Even the biggest tech companies do marketing wrong — and startups you haven’t heard of (yet) are doing it right. Demand Curve works with dozens of top companies coming out of Y Combinator, and a long list of others around the world. Asher King Abramson has seen it all, including the best practices going into the new year.
He’ll be critiquing a selection of efforts on stage, and you can have yours included if you submit your ad and landing page assets here.
Media historians, tech designers, science fiction writers, and activists on how and why “slow media” curated by humans should replace today's social media — My quest to imagine a different reality.
Jess Weatherbed / The Verge : Anthropic updates Claude to add custom writing styles tailored to users, aiming to personalize its replies,...
