Retrotechtacular: Häfla Hammerforge Healed

Visit any renaissance fair across the United States this fall and you’ll undoubtedly find a blacksmith. He’ll be sweating away in a tent, pounding on a piece of glowing steel set against an anvil. While the practice of the single blacksmith endures today, high-production ‘works of days past required increasing amounts of muscle. The more tireless the muscle, the better. The manual efforts of the blacksmith were replaced by huge hammers, and the blacksmith needed only to turn the piece between impressions and maintain a healthy respect for the awesome crushing power of the machine.

Last week, blacksmith enthusiasts completed restoration work on the Häfla hammer in Finspang, Sweden. The 333 year old hydraulic hammer hadn’t been used since 1924, when operations ceased at the Häfla Hammerforge. The ‘works was built in 1682 and used the German method of forging, which had been introduced to Sweden in the 1500s. Steel production was revolutionized in the 1800s by the Bessemer process, which resulted in a much stronger product. Continue reading “Retrotechtacular: Häfla Hammerforge Healed”

Retrotechtacular: The Hammond Novachord

Just when we thought we’d heard of all the cool early synthesizers, a tipster rattled our jar with news that someone completely restored a Novachord. These spinet piano-shaped prototypical synthesizers were made by Hammond for only four years. About a thousand of them were built before sales sagged and parts became scarce in 1942. It is estimated that only 200 or so are still around today.

The Novachord’s sounds are generated by a bank of twelve monostable vacuum tube oscillators. Each one is tuned to a pitch of the chromatic scale in what is called divide-down architecture. [Hammond] and his co-creators [John Hanert] and [C.N. Williams] used the property of dividing a frequency in half to generate the same tone, but one octave lower. This design means that all 72 notes can be played at the same time. Adjustable formant filters shape the often otherworldly sounds, which are then passed through flexible tube-based envelopes.

[Phil] knew it would be a big job to restore a Novachord in any condition. Thousands of passive components all had to be replaced. The cabinet bore all the hallmarks of a well-used parlor instrument—water rings from cocktails, scratches, and cigarette burns galore. [Phil] says that woodworking really isn’t his thing, but he did an outstanding job nonetheless of sanding every nook and cranny and applying several coats of stain. There are tons of drool-inducing pictures on his project site, and several clips of [Phil] really putting it through its paces.

Thanks for the tip, [Mike]!

Retrotechtacular is a weekly column featuring hacks, technology, and kitsch from ages of yore. Help keep it fresh by sending in your ideas for future installments.

Robots Are Coming For Our Jobs. Just Not All of Them.

There is a lot to be said for replacing certain kinds of jobs with robots. Most people would agree that replacing physical human labor with automation is a good thing. It’s especially good to automate the dangerous kinds of labor like some facets of factory work. What about automation in fields that require more mental labor, where physical strain isn’t the concern? Is replacing humans really the best course of action? A year ago, a video called Humans Need Not Apply set forth an explanation of how robots will inevitably replace us. But that narrative is a tough sell.

Whether it is even possible depends on the job being automated. It also depends on how far we are able to take technology, and the amount of labor we are willing to offload. Automation has been replacing human workers in assembly and manufacturing industries for years. Even with equipment and upkeep expenses, the tireless nature of robotic workers means dramatically lower overhead for businesses.

Many of the current forms of factory automation are rather dumb. When something goes wrong and their task is compromised, they keep chugging away. That costs time and money. But there are companies out there producing robots that are better on many levels.

May Your Robot Overlords Be Cute and Cuddly

baxter-heroIn 2013, Rethink Robotics started filling orders for a new line called Baxter. They are a class of general purpose robot that can be programmed to do many kinds of manual tasks. Baxter bots have vision, and they can learn how to do a job simply by watching. They don’t need to be programmed in the traditional sense.

Baxter even has a face – a screen that shows different expressions depending on his state. When he’s in the midst of a task, his eyes are cast downward. If something goes wrong, he stops what he’s doing. His cartoon face appears sort of shocked, then sad. He goes into safe mode and waits to be fixed.

Continue reading “Robots Are Coming For Our Jobs. Just Not All of Them.”

Retrotechtacular: One Does Not Simply String Up a Half-Million VDC Transmission Line

It takes strong and determined population to build a lasting civilization. If the civilization includes electricity and the inhabitants live in a hilly place with an often-unforgiving climate, the required strength and determination increases proportionally. Such is the case of the gentlemen who strung up the first half-million VDC transmission line across New Zealand, connecting the country’s two main islands.

Construction for the line known as the HVDC Inter-Island link began in 1961. It starts at the Benmore hydroelectric plant on the south island and runs north to Cook Strait via overhead cables. Then it travels 40km underwater to the north island and ends near Wellington. This is the kind of infrastructure project that required smaller, preliminary infrastructure projects. Hundreds of miles of New Zealand countryside had to be surveyed before breaking ground for the first tower support hole. In order to transport the materials and maintain the towers, some 270 miles of road were laid and ten bridges were built. Fifteen camps were set up to house the workers.

The country’s hilly terrain and high winds made the project even more challenging. But as you’ll see, these men were practically unfazed. They sent bundles of steel across steep canyons on zip lines and hand-walked wire haulage rope across gullies because they couldn’t otherwise do their job. Six of these men could erect a tower within a few hours, which the filmmakers prove with a cool time-lapse sequence.

Splicing the mile-long conductors is done with 100-ton compressors. Each connection is covered with steel sleeve that must be centered across the joint for optimum transmission. How did they check this? By taking a bunch of x-rays with a portable cesium-137 source.

Continue reading “Retrotechtacular: One Does Not Simply String Up a Half-Million VDC Transmission Line”

The Machine that Japed: Microsoft’s Humor-Emulating AI

Ten years ago, highbrow culture magazine The New Yorker started a contest. Each week, a cartoon with no caption is published in the back of the magazine. Readers are encouraged to submit an apt and hilarious caption that captures the magazine’s infamous wit. Editors select the top three entries to vie for reader votes and the prestige of having captioned a New Yorker cartoon.

The magazine receives about 5,000 submissions each week, which are scrutinized by cartoon editor [Bob Mankoff] and a parade of assistants that burn out after a year or two. But soon, [Mankoff]’s assistants may have their own assistant thanks to Microsoft researcher [Dafna Shahaf].

[Dafna Shahaf] heard [Mankoff] give a speech about the New Yorker cartoon archive a year or so ago, and it got her thinking about the possibilities of the vast collection with regard to artificial intelligence. The intricate nuances of humor and wordplay have long presented a special challenge to creators. [Shahaf] wondered, could computers begin to learn what makes a caption funny, given a big enough canon?

[Shahaf] threw ninety years worth of wry, one-panel humor at the system. Given this knowledge base, she trained it to choose funny captions for cartoons based on the jokes of similar cartoons. But in order to help [Mankoff] and his assistants choose among the entries, the AI must be able to rank the comedic value of jokes. And since computer vision software is made to decipher photos and not drawings, [Shahaf] and her team faced another task: assigning keywords to each cartoon. The team described each one in terms of its contextual anchors and subsequently its situational anomalies. For example, in the image above, the context keywords could be car dealership, car, customer, and salesman. Anomalies might include claws, fangs, and zoomorphic automobile.

The result is about the best that could be hoped for, if one was being realistic. All of the cartoon editors’ chosen winners showed up among the AI’s top 55.8%, which means the AI could ultimately help [Mankoff and Co.] weed out just under half of the truly bad entries. While [Mankoff] sees the study’s results as a positive thing, he’ll continue to hire assistants for the foreseeable future.

Humor-enabled AI may still be in its infancy, but the implications of the advancement are already great. To give personal assistants like Siri and Cortana a funny bone is to make them that much more human. But is that necessarily a good thing?

[via /.]

It’s Time to Roll Your Own Smartwatch

Giant wristwatches are so hot right now. This is a good thing, because it means they’re available at many price points. Aim just low enough on the scale and you can have a pre-constructed chassis for building your own smartwatch. That’s exactly what [benhur] did, combining a GY-87 10-DOF module, an I²C OLED display, and an Arduino Pro Mini.

The watch uses one button to cycle through its different modes. Date and time are up first, naturally. The next screen shows the current temperature, altitude, and barometric pressure. Compass mode is after that, and then a readout showing your step count and kilocalories burned.

In previous iterations, the watch communicated over Bluetooth to Windows Phone, but it drew too much power. With each new hardware rev, [benhur] made significant strides in battery life, going from one hour to fourteen to a full twenty-fours.

Take the full tour of [benhur]’s smartwatch after the break. He’s open to ideas for the next generation, so share your insight with him in the comments. We’d like to see some kind of feedback system that tells us when we’ve been pounding away at the Model M for too long.  Continue reading “It’s Time to Roll Your Own Smartwatch”

Retrotechtacular: Robots, Robots Everywhere, with Kitschy Pronunciation

One of the great things about the human intellect is that we have the ability to build machines of varying complexity to do our bidding. As a major proponent of technology, the Chevrolet automobile corporation once dreamed of a future where the American housewife’s most mundane tasks are handled with the push of a button—one that sets a robot butler into action.

Chevy shows us what this future might look like in this short film, which they presented at the 1940 World’s Fair. A housewife’s faithful ‘robot’, pronounced throughout the picture as ‘robe-it’, has gone on the fritz. Naturally, she calls for a repairman. We see from the console controller that Roll-Oh the Robe-it can take care of all kinds of housewifely duties: he can answer the door and the phone, wash dishes, clean house, make beds, fetch hats, get dinner, and fix the furnace (and only the furnace). And that SCRAM! function? That’s never explained. We like to think it has to do with getting kids off the lawn, or could be used in conjunction with ‘get door’ to chase away would-be burglars. We get a glimpse of this when Roll-Oh answers the door and scares the daylights out of a young [Gary Sinise*] delivering flowers in a cop uniform.

Roll-Oh’s upper limbs have several Swiss Army knife-like implements in them. He uses a sharp one to cut the ribbon off of the flower box. Upon seeing the flowers, he gives them a gentle misting with his sprayer attachment. Dropped petals are no problem for Roll-Oh. He promptly vacuums them up from the thin industrial sound stage carpet with his big metal feet. Roll-Oh is then tasked with getting dinner. This amounts to him painstakingly opening a couple of cans and lighting candles with the torch hidden in his face.

While Roll-Oh the large ductwork butler is only a dream, Chevy wants you to know that smaller robe-its are all around us already. They’re regulating the heat in our stoves, browning our bread without burning it, and brewing our coffee in cool double-globe glass percolators. These tiny servants are capable of performing other tasks, like shutting off machinery when humans are too close, or sensing heat and engaging fire suppression systems. There is brief mention of something called the Petomat, an automatic dog feeding system which is essentially a bowl of food hidden in a latched box. The latch opens rather violently when the alarm clock connected to it goes off.

Robe-its are also performing more serious tasks, like keeping airplanes level and headed in the right direction. Of course, they’re also abundant in Chevrolet automobiles. A small one in the carburetor administers the proper mix of “gasoline calories and fresh air vitamins” to the engine. It’s rare to get to this level of technical detail, you know. Others watch over the spark, the intake manifold, and the voltage regulation. Up in the cab, friendly robe-its will happily traverse the AM dial at the push of a pre-set.

*Probably not actually [Gary Sinise].

Continue reading “Retrotechtacular: Robots, Robots Everywhere, with Kitschy Pronunciation”