95
points
gene-h
Joined 1,608 karma
- gene-h parentThis project was cancelled a couple months ago: https://breakingdefense.com/2025/06/darpas-draco-nuclear-pro...
- There are proposals for the 6G standard to support Integrated Sensing and Communication(ISAC)[0]. So the hardware might natively be able to support gait recognition. The use cases given are UAV detection and localization. It sort of seems like this could bring Vernor Vinge's localizer mesh to reality, privacy implications be damned [0]https://www.ericsson.com/en/blog/2024/6/integrated-sensing-a...
- Industrial robots at least are very reliable, MTBF is often upwards of 100,000 hours[0]. Industrial robots are optimized to be as reliable as possible because the longer they last and less often they need to be fixed, the more profitable they are. In fact, German and Japanese companies came to dominate the industrial robotics market because they focused on reliability. They developed rotary electric actuators that were more reliable. Cincinnati Millicron(US) was out competed in the industrial robot market because although their hydraulic robots were strong, they were less reliable.
I am personally a bit skeptical of anthropormophic hands achieving similarly high reliability. There's just too many small parts that need to withstand high forces.
[0]https://robotsdoneright.com/Articles/what-are-the-different-...
- And why wouldn't it work? Linear slide like mechanisms consisting of a silver surface and single molecule have been demonstrated[0]. The molecule only moved along rows of the silver surface. It was demonstrated to stay in one of these grooves up to 150 nm. A huge distance at this scale.
- A similar single motor robot hand has been made that uses electrostatic clutches instead of mechanical clutches[O].
- At some point, making mechanical watches more complicated will require going digital. It is possible to make very small gears with semiconductor processes, however, very small gears wear out fast due to stiction.
In order for gears to work they must have sliding contact and that means wear. Mechanisms based on flexures don't have this problem, but this requires building the clock very differently. It might be possible to implement many of these complications using flexure based logic[0].
- 11 points
- Perovskite solar cells could be and dye sensitized solar cells can be[0]. The better question is why should one make solar cells at home?
[0]https://www.instructables.com/How-to-Build-Use-A-Dye-Sensiti...
- It's useful for grid storage. Very large amounts of hydrogen are already stored in salt domes[0]. Current salt domes have volumes in the range of hundreds of cubic kilometers and can support pressures around 50-150 bar, translating into storage of thousands of tons of hydrogen. Along the texas gulf coast, there are hydrogen storage facilities that each store enough hydrogen to translate to around 100 GWh chemical energy. Being able to convert that chemical energy with 40% end to end efficiency means one site could store 40 GWh. In comparison, in 2023 the entire world had only around 56-200 GWh of battery storage capacity[1] installed.
[0]https://energnet.eu/wp-content/uploads/2021/02/3-Hevin-Under... [1]https://www.rethinkx.com/blog/where-is-all-the-battery-stora...
- 1 point
- You could use it to model a CNC machine, but you'd probably need to make new components. It would also be difficult to model the changing load as the spinning tool plunges into metal, but you could model dynamic deformation and response of the machine given the loads. It is difficult to model changing 3D contacts in modelica, although a new approach to solving newton's equations, dialectic mechanics may fix this problem
NEMA isn't a type of motor, it's a connection standard used for motors. You might be able to model stepper motors with some of the components in the modelica standard library for magnetic modeling.
- Europa Clipper also used a new approach for designing spacecraft. It's NASA's first major spacecraft designed with Model Based Systems Engineering(MBSE)[0]. Using diagrams in SysML to keep track of power use and interfaces is supposedly better than using spreadsheets
- There was a whole field called fluidics that focused on performing operations analogous to those performed with electricity. This[0] gives a good overview of how fluidics worked. One the most commonly used elements of fluidics was the fluidic amplifier. This worked by using sideways flow of lower pressure to redirect a more powerful jet of fluid between two ports. Fluidic amplifiers could be made with frequency responses in the KHz range, so they have been used to amplify sound[1].
There was some brief interest in fluidics because it was cheap, more reliable than electronics of the time, and could function in extreme environments. So it found use in industrial automation and aircraft systems. This document[2] from NASA shows some applications it found at the time. Univac even built a completely fluidic 4 bit digital computer[3].
Fluidics is still around today, but used for very niche applications. One of the most absurd uses I've heard of was getting around rules prohibiting active aerodynamics in Formula 1. In 2010, the McLaren devised a system where the driver could cover a hole on part of the car, causing flow to be redirected to a fluidic amplifier that redirected flow over the rear spoilers causing them stall, allowing drag to be reduced on the straight aways. IIRC the entire purpose of the system was to get around rules which prohibited doing this with moving parts.
[0]https://miriam-english.org/files/fluidics/FluidControlDevice...
[1]https://acoustics.org/pressroom/httpdocs/132nd/2aaa8.html
[2]https://ntrs.nasa.gov/api/citations/19730002533/downloads/19...
[3]https://dl.acm.org/doi/pdf/10.1145/1464052.1464112
[4]https://us.motorsport.com/f1/news/banned-technical-analysis-...
- Atomically precise manufacturing, that is tech for building structures atom by atom. This is starting to be done regularly in labs. Machines for making single atom transistors are a commercial product now[0]. Forming covalent bonds at desired locations through positional control of reactants has been demonstrated[1]. This is potentially more scalable than the aforementioned approach. ML seems to be enabling too[2].
[0]https://www.zyvexlabs.com/apm/products/zyvex-litho-1/
- What's more important is that they demonstrated making diamond at 1 atm and lower temperatures(1025 C). This is compared to ~50,000 atm and ~1500 °C diamond is conventionally made at. The diamonds they made were very small, but this is a new process and optimization might enable it to make bigger diamonds.
- The efficiency demonstrated is a joke. There are better ways and more direct ways to extract power from humidity gradients[0]. In a dry environment, the maximum amount of energy extractable per unit water is fairly high, about the energy density of lead acid batteries[1]. There are even proposed large scale power plants for generating power from humidity gradients[2].
The catch? It needs fresh water and works best in hot dry areas where freshwater is at a premium. Also, the powerplant would make surrounding communities up to 100 km extremely humid. Trading freshwater for electricity is not a very attractive proposition...
[0]https://www.researchgate.net/publication/279853268_Scaling_u...
[1]https://www.researchgate.net/publication/237493406_A_Dunking...
- The IM-1 Lander was supposed to land using a LIDAR altimeter, but they forgot to remove the safety before launch. They tried to make a last minute software change to use an experimental navigation system from NASA to get altimetry, but this didn't work. So the lander landed using visual navigation and IMU data only for the last 15km to the surface.
It probably would have landed upright if the LIDAR worked. It is impressive that it landed as intact as it did
[0] https://arstechnica.com/space/2024/02/it-turns-out-that-odys...
- Much of the damage to the power grid can mostly be mitigated by turning off electricity, although this is a difficult thing for power companies and grid operators to do. One issue this article doesn't discuss is the risk to undersea internet cables[0]. Undersea fiberoptic cables need repeaters, these need electricity, so they have very long conductors and it's expected that sea water's conductivity could make induced currents worse. Shutting off power won't necessarily work, because induced currents could be 100x more than the equipment is rated for. Although, global connectivity is still likely to exist.
- There has been more interesting work on using transformers for robot motion planning[0]. Getting a robot arm from point a to b without hitting stuff is a very difficult problem, the problem is both high dimensional and continuous. Previous planning methods are both computationally intensive and not very good. This is one reason why robot motion appears 'unnatural' and robots generally being bad at many tasks we'd like them to do. This approach appears pretty competitive with other planning methods, planning near optimal paths faster.
- NIF uses inefficient lasers because they were cheap to build at the time and because NIF is a science experiment. Lasers have gotten better. And lasers aren't the only way ICF fusion could be carried out, it may be possible to use ion beams instead[0].
It does not matter if fusion reactions last microseconds if they generate more energy. Using optimistic, but not unrealistic assumptions, it appears feasible for electricity costs to reach $25 per MWh[1] with ICF. With the most important factors driving cost being achieving high gain and yield per shot
[0]https://ieeexplore.ieee.org/document/650904/
[1]https://royalsocietypublishing.org/doi/10.1098/rsta.2020.005...
- The work from Rochester mentioned briefly is probably the second best performing fusion experiment and this was done with a laser that's at least 29 years old.
What the article doesn't mention is that the approach demonstrated could decrease the cost of fuel pellets. In NIF, the laser energy is converted to X-rays which compress the fuel pellet using expensive elements like gold. What they demonstrated is that the fuel pellet can be imploded directly with the lasers.
- >There is some radiation with some designs, but the basic idea of fusing hydrogen into helium works without radioactivity. Per unit energy D-T fusion, the fusion reaction most likely to be viable near term, generates more neutrons than the fission of U235. The U235 fission reaction[0] releases 3 neutrons and 202.5 MeV of energy, 8.8 MeV escapes as antineutrinos, so let's call it 193.7 MeV, so 64.5 MeV per neutron.
D-T fusion generates 17.6 MeV and 1 neutron, so 17.6 MeV per neutron. So for the same amount of energy produced it makes about 3.5x more neutrons. Worse yet, in D-T fusion, 14.1 MeV of the energy released is in the kinetic energy of the neutron. Compare this to the 8.8 MeV energy of all three neutrons in a fission reaction. Neutron activation and damage are significant engineering concerns for fusion reactors.
[0]https://en.wikipedia.org/wiki/Uranium-235 [1]https://en.wikipedia.org/wiki/Deuterium%E2%80%93tritium_fusi... https://en.wikipedia.org/wiki/Uranium-235