Manufacturing Drones - The Miracle We Need To Automate And Commoditize Manufacturing

Summary:

The U.S. is at the precipice of losing AI due to being a decade or more behind on energy, manufacturing, and physical scaling. China has more AI engineers, the lowest cost compute, no longer needs U.S. designed chips, and has the manufacturing capacity to scale and monetize AI in the physical world. Six axis manufacturing drones can level the playing field, re-commoditizing manufacturing and being the miracle the U.S. needs to catch up, but it is a ~3 year race to scale. Six axis manufacturing drones leap far beyond humanoid robots as they out perform them on every major metric. Manufacturing is a $15 trillion/year market, but it enables everything else, including AI and energy. As Marx said, they who own the means of production rule the world.

Within a factory, six axis manufacturing drones can connect every production machine and every parts store with an AI enabled adaptive high speed logistics network that ends in a dense high speed assembly swarm. One dimensional production lines are replaced with a three dimensional matrix of production cells operating at a much higher clock speed. Six axis small and large drones alike, with integrated automatic tool and arm changing and recharging, operate in a hive of activity.

This technology will scale incredibly rapidly, disrupting existing manufacturing and supply chains as autonomous fast response manufacturing becomes available to almost every industry at low-cost. This will move as fast if not faster than AI, and even more revenue is at stake. We are seeking first adopters, open technology licensing, and supporting advisory roles.

Concept:

The general concept and capabilities of six axis drones is described in more detail here.

Opportunity

Manufacturing, along with materials, is ideally where to start in order to have truly global revolutionary impact. The scaling rates and cost reduction curves of almost all new technologies are limited by manufacturing. Automating, democratizing, and reducing the cost of manufacturing, so that it becomes labor cost insensitive, is technologically enabling for almost everything, including energy and AI. There are well over 100 different humanoid robot companies and a market size of around $5 trillion per year is expected by 2050. However, the world and AI will be a very different place by then and the need for physical automation might be orders of magnitude higher while the need for robots of specific human form will be only a small proportion of this. Tesla is basing ~80% of its future valuation on the Teslabot, although China has already surpassed Tesla with humanoid robots, just as it has with electric car revenues. 2027 to 2030 might be a more reasonable estimate for the dramatic scaling of AI and AI robotics, beyond which it is difficult to make reasonable predictions. Noting that the physical manifestation of AI in the real world via robotics, where it might disrupt the majority of global GDP, might be worth almost an order of magnitude more than pure AI.

Manufacturing is at the heart of civilization, it is the foundation of our high technology lives. Tool use is what separates us from our distant ancestors. Manufacturing enabled us to increase our per capita energy use ~100x and it provided us with economic prosperity. Six axis manufacturing drones are the next giant leap forward in manufacturing, comparable to Henry Ford’s creation of the moving assembly line in 1913. Humanoid robots are positioned to replace people on the production line, enabling automation of traditional labor roles. Six axis manufacturing drones can go far beyond what both people and humanoid robots can achieve, as they add the power of high speed acceleration and flight. Greatly speeding up assembly processes and also allowing for the utilization of the third dimension for both logistics and assembly access. Enabling a three dimensional matrix of adaptable work cells. Six axis manufacturing drones of many different sizes can also act at once, not constrained by person sized access pathways. Small six axis manufacturing drones might perform fine work, inserting fasteners, using power tools, operating inside difficult to access spaces, and so forth, while heavy lift six axis manufacturing drones might quickly move large assemblies that may weigh multiple metric tons. In some scenarios assemblies may never even touch the ground, being assembled in midair as they fly through a factory, providing fast access to all sides at once while avoiding fixed ground infrastructure. Factories can at once be both much smaller, as more manufacturing can be performed at higher speed in a much smaller and more three dimensional space, and also much larger, as distances can be traversed much more quickly, reducing the time cost of distance between machines and operations. This will enable even greater economies of scale as factories will be able to produce more and more complex products. Many different products, perhaps sharing similar processes, might be produced at high speed at once in the same factory.

An interesting aspect of manufacturing drones is that they might operate entirely indoors, although they may want to interface fairly directly with outdoor logistics. This can mitigate FAA constraints and avoid the need for certification, which can greatly accelerate the pace of development and deployment. Learning curves and scaling rates can be very fast, faster than many other six axis drone applications. Indoor operation also avoids visual sight and sound pollution that can impede operating consent. On the downside, indoor operation can make for a very windy workplace, especially if heavy lift drones are used. Six axis drones can directly mitigate the impact of turbulence on drone flying precision, but the workplace will need to be wind resistant. Drones, even if using shrouded propellers, might also be less safe around people if the manufacturing space is shared with people, which it probably should not be. If human operations are needed then a separated human safe environment might be used with drones transferring components between the two spaces.

Design:

Many different types of six axis drone might be used within a manufacturing context, the design constraints being somewhat different to outdoor operation. For example, flight distances and speeds might generally be less such that aerodynamics are of less concern. Acceleration and agility might be of greater concern due to the need to minimise travel time and optimize trajectories between operations, with thrust magnitudes in desired directions being optimized accordingly. A wide range of drone size will also be desired so as to match the wide range of operations that will be needed, from small manufacturing drones that might be used to install fasteners and small components, and perform functions typical of portable battery operated power tools, to larger drones that might be used to transport and place larger components. Heavy lift drones might also be desired for transporting large assemblies, thereby reducing the need for fixed infrastructure such as moving assembly lines, although the much greater downwash from heavy lift drones might limit their use. Some six axis manufacturing drones might include reversible motor controllers and propellers that enable flight at more extreme orientations, for example, inverted flight, for performing some more difficult tasks.

Tools and arms:

Six axis manufacturing drones differ substantially from humanoid robots in that they can serve directly as a six axis platform for tools. Performing functions characteristic of a six axis robotic arm, but at significantly lower cost and higher longevity (perhaps only six moving parts), without requiring a fixed base and without being limited in reach. The six axis manufacturing drone can automatically change tools using an interface that can include power and communications and might consist of a standardised tapered socket. For example, they might directly interface with a wrench, hammer, orbital sander, welder, electric screw driver, and so forth, and position and orient it, in all six degrees of freedom to perform a given task. Force control is also possible and drones might automatically swap between tools as needed. With standardized tool socket interfaces various extenders might be used, for example, a high speed delta robot type extender that could actively compensate for drone movement and enable sub millimeter tool positioning. Robotic arms would likely directly connect to these standardized tool socket interfaces - a six axis manufacturing drone might have multiple tool sockets interfaces in different locations, which would allow for their automatic changing. In addition to allowing six axis manufacturing drones to quickly attach the right tool or robotic arm for a given task, automatic tool changing also allows for the independent development and maintenance of tools and robotic arms. Tools and robotic arms often entail wear components that might need frequent replacing, for example, cutting surfaces that need sharpening or replacing or actuators that have limited longevity. By separating the platform from the tool, maintenance requirements can be optimized separately. A six axis manufacturing drone can quickly replace a tool and a separate process for maintaining tools can be established - perhaps using other six axis manufacturing drones to maintain tools. Six axis manufacturing drones enable a full manufacturing system that is potentially able to manufacture and repair manufacturing drones, dedicated manufacturing equipment and machinery, and the tools that they all use. This can be a true automated manufacturing revolution.

Batteries:

Just as six axis manufacturing drones can precisely position themselves for automatic tool changing they can also precisely position themselves for automatic recharging and/or battery swapping. For example, they can directly land on a charging electrical connector or a battery. In the case of battery swapping multiple batteries or an internal small high power battery are needed so that the drone maintains power throughout the battery swapping process. High power high cycle life batteries are generally desired and while high specific energy is also desirable, with close and frequent access to charging stations this is not essential. Lithium iron phosphate batteries will generally suffice, although CATL’s Naxtra sodium battery, which is claiming greater than 10,000 cycles, 175Wh/kg, and up to 12C charging rates (5 minute charging), seems like an even better prospect. They might be capable of less than $0.01/kWh/cycle, which is a critical cost metric that is primarily driven by very high cycle life. Battery swapping enables greater utilization rates of the drone which can be useful in some scenarios, including high power applications, like say welding drones. This depends on the ratio of flight endurance time to recharge time. With battery swapping other drones can also deliver charged batteries to the desired location and remove discharged batteries for recharging. Taken to extremes this can even occur in midflight - six axis manufacturing drones have the flying precision needed to dock with another drone and swap their batteries while they are flying. In some scenarios involving very localized operation six axis manufacturing drones might also operate on a short power tether, allowing for continuous operation. The six axis manufacturing drone might land on the tether connector, similar to how it might land on a charging connector or new battery, and operate from that tether as needed. Long tethers can present a risk to other flying drones so some care is needed to keep the tether out of the way.

Sensors:

High precision flight necessitates sensor accuracy that exceeds flying precision, and this includes orientation. Millimeter level position accuracy and higher might generally be needed. Standard low-cost solid state IMUs that include three axis accelerometers, gyros, and magnetometers will help, but absolute position and orientation sensing is needed. High acceleration platforms will make gravity vector estimation more challenging and factory environments might also make magnetometers that detect the Earth's magnetic field problematic. The equivalent of indoor GPS is needed, and likely multiple displaced units within the drone so as to enable the absolute sensing of orientation. Ultrasound is one possibility although dense drone swarms operating in high noise environments may present problems. Ultra wide band positioning systems for indoor use are perhaps the most likely prospect, but this might need to be augmented with localised relative sensing such as lidar, radar, and camera systems. Higher precision work that might utilize an actuated end effector to achieve sub millimeter accuracy would likely need to depend on some kind of local imaging system, for example, cameras. Sensors also need to be flight weight and cost compatible, which can be challenging for smaller drones. A factory is a highly structured environment and there are technical solutions to most of these problems, as the market for six axis manufacturing drones increases it would be expected that much higher accuracy low-cost sensor systems would be developed.

Impact:

Six axis manufacturing drones enable the generalized automation and commoditization of manufacturing at small and large scales. Factories integrating CNC machinery and other heavy machinery with six axis manufacturing drones might become like AI enabled universal replicators, with the ability to mass produce a large range of complex products on demand and at low-cost. Especially when combined with external much faster and lower-cost six axis drone logistics systems for materials supply and product delivery. Not only does this disrupt the ~$15 trillion dollar per year manufacturing industry, but it also enables democratization and distributed operation, and dramatic cost reduction and scaling of manufacturing in support of a future world of abundance. It is a big step towards AI enabled manufacturing self-sufficiency. This enables the much faster scaling of buildings, energy systems, transport systems, food production systems, materials production and recycling systems, industrial plants, nature conservation systems, general innovation, and so forth. Greatly reducing the cost of living while increasing living standards and conserving nature. It also enables much faster scaling of AI and the direct monetization of AI at much higher revenues than pure AI alone, and in ways that less compromise the natural and human world. AI enables six axis manufacturing drones and six axis manufacturing drones are in turn critical to the future of AI. As a consequence, six axis manufacturing drones warrant investment even in excess of current AI investment.

Next steps: