Dr. David Pares describes the Variable Electromagnetic Drive as an experiment born from thunderstorms, fractal antennas, crossed electromagnetic fields, and a quest for propulsion based on natural phenomena. In his latest APEC presentation, Pares showcases AI-assisted software is now helping him model field geometry, configure multi-engine arrays, explore fractal iterations, and automate the startup process of a drive he believes can compress spacetime using electromagnetic structure. Alongside that update, Reid Sherman presented a separate but related milestone: a year-long replication effort that he says produced repeatable directional movement in his own lower-power VEM test setup.

APEC’s Two-Part VEM Drive Update

The presentations by David Pares and Reid Sherman formed a natural two-part update on the VEM Drive, the unconventional propulsion concept Pares has been developing through his startup, Quantum Electro Dynamics R&D. Sherman’s presentation focused on replication: whether a researcher outside Pares’s own lab could reproduce directional movement with a VEM-style apparatus. Pares’s presentation focused on modeling and scale: how AI tools, new software, multi-engine arrays, and a larger demonstrator might move the project into its next phase.

That pairing gave the APEC session a useful structure. Sherman opened as the practical experimental voice, describing a year spent trying to test the VEM Drive concept for himself. He introduced himself as a NASA Nebraska student fellow working out of the University of Nebraska–Lincoln, and framed his work around a simple aerospace problem: rockets are powerful and reliable, but their dependence on propellant makes deep-space missions costly and limited. For Sherman, the VEM Drive is interesting because it claims to explore propulsion without conventional reaction mass.

Pares followed with the broader origin story and engineering roadmap. He described himself as a meteorologist, physicist, astronomer, teacher, and engineer-scientist who approaches the work from an empirical background. Rather than beginning with a finished theory, he said he has spent years observing, building, testing, modifying, and trying to understand what the experiments appear to show. His presentation mixed physics language, weather history, lab experience, software development, and long-range spaceflight goals.

The result was less a finished declaration than a project report. Pares and Sherman presented the VEM Drive as a living research program, not a settled propulsion technology. Their claims remain outside mainstream aerospace theory, but the talks also made clear that both presenters see the next step as better testing, stronger replication, more precise measurement, and improved tools for modeling the field effects they believe they are seeing.

The New Hook: AI Enters the VEM Program

The freshest part of Pares’s update was his claim that AI has changed the pace and structure of the VEM Drive project. He said he had been doing extensive work with Claude AI and had written nine software packages connected to the drive. For a project that has often been described through physical demonstrations, videos, and hand-built test articles, the new emphasis on software gave the presentation a different feel.

Pares described the AI-assisted tools as a way to visualize and engineer VEM field structures. He said the software can reproduce field patterns, display fractal designs, vary the number of fractal iterations, and help configure engines into quad, hexagonal, octagonal, and larger arrangements. He also described tools that allow the user to roll, pitch, yaw, and adjust engines within these arrangements, then interrogate field overlap points to examine their behavior.

That is a meaningful shift in how the project is presented. For years, Pares has described the VEM Drive as an empirical effort: build an engine, power it, observe its effect, adjust the geometry, repeat. In the AI modeling phase, he is now describing something closer to a design environment — a way to take accumulated experimental intuition and organize it into interactive tools that can guide new builds.

The software does not, by itself, prove Pares’s interpretation of the VEM Drive. But it does give the project a more coherent engineering language. Instead of relying only on test videos and verbal descriptions, Pares can now show field shapes, engine configurations, fractal iterations, and overlap regions in ways that are easier for others to discuss, challenge, and potentially reproduce.

A Drive Inspired by Thunderstorms

Pares’s explanation of the VEM Drive begins with weather. He said his thinking grew from atmospheric physics, especially thunderstorms, ionization, freezing-level frequencies, crossed fields, and unusual pilot reports. In his view, some high-energy storm systems may create local electromagnetic conditions that produce effects not normally associated with propulsion.

This origin story is central to understanding how Pares presents the VEM Drive. He is not describing it as a conventional electric thruster, ion engine, or RF antenna system. Instead, he describes it as an attempt to reproduce, in a controlled laboratory setting, an effect he believes nature may already produce under rare conditions. The thunderstorm is the template; the VEM engine is the artificial mechanism.

Pares talked about charge structures in clouds, frequencies associated with storm regions, and the freezing level as especially important to the frequencies he has tried to mirror in his fractal designs. He described thunderstorms as environments where crossed fields can arise naturally, and he connected those ideas to pilot reports of unusual linear displacement. In his telling, those reports became the clues that led him to search for a laboratory analogue.

This is where the VEM Drive clearly diverges from mainstream propulsion theory. Conventional aerospace physics does not currently recognize thunderstorm-inspired crossed electromagnetic fields as a demonstrated route to propellantless thrust or spacetime compression. But Pares’s presentation was not framed as a textbook derivation. It was framed as an empirical journey: nature seemed to suggest a phenomenon, and he has spent years trying to recreate and engineer it.

The Hardware: Fractals, Fields, and RF Power

The VEM Drive hardware is visually distinctive. Pares and Sherman both described fractal structures arranged to reproduce crossed-field behavior. Pares identified these as modified Koch-curve fractals, arranged in tripole fields and excited with RF energy. To an outside observer, the components may look like unusual antennas, but Pares argues that their function is more specific than ordinary radiation.

According to Pares, the fractal geometry matters because it compresses the behavior of a much longer antenna into a smaller structure while creating nested current loops. He said those loops are central to the drive’s behavior. In his explanation, the iterations of the fractal are not decorative or incidental; they are part of what shapes the field and produces the longitudinal compression he believes is responsible for the pulling effect.

The engine operates through a power chain involving 48-volt systems, linear amplifiers, tuners, coaxial cables, and Faraday-cage testing. Pares said the drive is operated in continuous-wave mode because the field must be built and maintained. He described the startup process as sensitive to material, geometry, fractal design, and tuning, and said it cannot simply be switched on instantly like a conventional appliance.

He also emphasized that different materials behave differently. Aluminum, copper, silver, and layered material concepts all came up in the discussion. Pares said thickness affects how current concentrates in the fractal iterations, while material choice affects how the engine initiates and sustains its field. These details gave the presentation an engineering texture: the VEM Drive, as Pares describes it, is not one simple device but a system of geometry, material, frequency, field structure, and startup behavior.

Reid Sherman’s Replication Effort

Sherman’s presentation provided the second major thread of the story. His role was not to explain the entire VEM Drive theory or to claim that every implication had been proven. He said his job was to ask a simpler and more important experimental question: can the reported effect be reproduced independently?

To frame that question, Sherman showed a QED Model 19 test video from December 15, 2023. He described the test as a 1,500-watt run over a 30-second interval, with the VEM system suspended and a dial moving as the device powered up. Sherman said QED had reported a pulling force of more than 15 newtons from the system, a figure he compared to the thrust of NASA’s X3 Hall thruster.

Sherman’s reaction to that video was openly enthusiastic. He described the apparent strength of the system as “insane” and used the comparison to explain why the VEM Drive seemed worth testing. At the same time, he framed the video as motivation for replication rather than as the final scientific word. If the claim is to become a reliable engineering system, he said, it has to be repeatable.

After roughly a year of work, Sherman reported that he was able to reproduce repeatable directional movement using his own VEM Drive test setup. His biggest milestone, he said, came on January 11, when he completed 10 successful tests in a row. For Sherman, that sequence mattered because it suggested the effect was not simply a one-time anomaly or “catching lightning in a bottle,” but something that could be generated repeatedly under similar conditions.

What Sherman’s Tests Showed — and What Comes Next

Sherman showed several test videos from his own work. One early video used a weak nylon-string mount and produced rotation around the mounting mechanism. Sherman described that setup as imperfect, but still important because it was his first successful test showing movement generated by the system.

Another video showed one of the 10 successful tests. Sherman noted that the dial moved from right to left rather than straight down because the focal point was not perfectly aligned. That detail was useful because it showed both the claimed directionality of the effect and the practical difficulty of controlling it. In Sherman’s interpretation, the setup was producing movement, but not always in the perfectly intended direction.

One of the most striking moments came when Sherman described a camera freezing during a powered run. He said the expanded view froze when power went into the system, apparently because of electromagnetic energy in the air, even with a Faraday cage in place. When power was cut, the camera view returned. Sherman did not present that as a polished result; he presented it as part of the reality of testing a high-RF experimental system.

His conclusion was careful but positive. He said the most important point was that movement had been repeatedly generated under similar test conditions. He also said there is still a lot of work to do. His planned next steps include a larger and more stable test rig, stronger PCB fractal arrays, and a multipoint interferometer setup so he can measure the field at different locations.

The Videos as Part of the Experimental Record

The VEM Drive story is closely tied to videos. QED has used videos to document scale movement, suspended-engine tests, enclosed-object experiments, and optical effects. Sherman’s presentation leaned into that record by showing QED’s Model 19 footage and then adding his own replication videos to the discussion.

For a supportive but responsible feature, the videos should be treated as part of the experimental record rather than as final proof. They are central to why Pares, Sherman, and QED believe the phenomenon deserves continued attention. They show what the team says it is observing and what motivates the next round of tests.

Pares also discussed earlier optical experiments involving lasers, mirrors, springs, graph paper, and smoke visualization. In one account, he described sending pen lasers through a field and seeing apparent bending around what he interprets as a warp field. He also acknowledged that some of those early observations were not quantified at the time, which gives the story a useful balance: the experiments were suggestive to him, but the measurement program is still developing.

That balance is important. The videos have not settled the VEM Drive for mainstream science, but they also remain the reason the project continues to attract attention. Pares and Sherman are not presenting the VEM Drive as a theory detached from hardware. They are presenting it as a video-documented, measurement-hungry experimental program that they believe should be tested more deeply.

Redshift, Blueshift, and Latency

Pares described several claimed signatures that he believes support the VEM Drive interpretation. One is redshift in the engine itself, which he presents as evidence of compression in the field. Another is blueshift behind the engine, which he associates with the rearward side of the effect. Together, he describes these as part of a pull-push structure.

He also emphasized latency — continued movement after power is shut off. Pares described a 300-watt run with a one-foot motor that reportedly produced about 10 grams of pull and then continued pulling for roughly 15 seconds after power was off. He interprets that as the field rebounding, dissipating, or returning to ground after the active input stops.

Sherman added to that discussion from his own perspective. He said that, in one of the experiments, the needle continued moving after power was cut, and he argued that the observed delay seemed too long to be explained simply by an inductor or capacitor in the system. He did not claim that this settled the explanation. Instead, he framed it as a question for future experimentation.

This may become one of the most important measurement targets in the next phase of the VEM Drive program. If Pares’s team can correlate force, optical behavior, field decay, and power-down timing in a controlled way, latency could become a clearer experimental signature. For now, it remains one of the effects Pares and Sherman find most intriguing and want to study more precisely.

AI Modeling and Multi-Engine Arrays

One of the most ambitious claims in Pares’s presentation involved multi-engine configurations. He said the new software can arrange VEM engines into quad, hexagonal, octagonal, and larger groupings. In those configurations, the fields can be overlapped and examined as a combined structure.

Pares described this overlap as coherent constructive interference, or CCI. In his model, overlapping fields from multiple engines can magnify the pulling effect. When asked during the Q&A whether the increase would be linear or exponential, Pares said it was exponential. He also said the software can probe points in the overlap region and generate curves showing the field behavior.

This is a key part of the project’s scaling logic. A single engine may show one level of force, but Pares believes multiple engines can be arranged so their fields reinforce each other. That is why the quad configuration matters so much in his roadmap. The software is not only a visualization tool; it is being used to plan how separate engines might work together.

The article’s title — “AI Modeling & Replication Experiments” — is strongest here. Pares’s AI-assisted modeling gives the project a claimed pathway toward scale, while Sherman’s replication experiments give it a claimed pathway toward independent repeatability. The two threads are different but complementary: one asks how to design larger systems, while the other asks how to reproduce the basic effect cleanly.

Geometry, Fractals, and the Sweet Spot

Geometry came up again and again in the APEC discussion. Pares said he had tested V-angle configurations from 50 degrees to 75 degrees and had converged around a roughly 70.5-degree geometry. When asked whether force rose to a peak around that angle and then fell off, he said it dropped on either side.

That angle dependence is important because it gives the VEM Drive story an engineering anchor. Pares is not merely saying that more power produces more force. He is saying that geometry, field orientation, material, fractal iteration, and startup conditions all interact. In his view, the VEM Drive works because the structure is tuned to create a specific field behavior.

The fractal discussion pushed this further. Pares said the modified Koch curves reduce the physical size of the antenna-like structure while preserving important behavior, and that the iterations create nested current loops. He connected those loops to longitudinal waveforms and to the compression effect he believes the drive produces.

The Q&A also moved into future designs. Jarod Yates asked about a three-dimensional fractenna, and Pares said he had already designed a cascading 3D fractal on paper with 18 iterations. He did not present it as a completed engine, but the exchange showed how the VEM discussion naturally moves from current tests into next-generation geometry.

Startup, Control, and the Initiator

Pares repeatedly emphasized that the VEM Drive has to be started correctly. In his description, the field must be initiated, built, and maintained. The process depends on the engine type, material, fractal structure, and operating conditions. He said he has historically done much of this manually, with a success rate that improved as he better understood how long the field needed to build.

That is why the “Initiator” software matters. Pares said QED is developing AI software that would connect directly to the engines, start them, build the field, and monitor the characteristics that are currently handled by hand. If the drive’s operation depends on a delicate startup sequence, then automation could be essential for repeatability.

The idea also connects to replication. A system that requires manual tuning by its inventor is harder for outside researchers to evaluate. A system with documented startup parameters and automated control routines could become easier to reproduce. In that sense, the Initiator is not only a convenience. It could become part of the project’s experimental discipline.

Pares presented this as a major advancement. The VEM Drive, in his description, is becoming less dependent on isolated intuition and more dependent on software-guided operation. Whether the broader claims ultimately hold or not, that shift could make the work easier to test, compare, and improve.

Lab Realities and Engineering Constraints

The VEM Drive story also has a practical, human side. Pares described QED’s lab as a modified garage with a 3-by-5-foot Faraday cage. That limited space has shaped what the team can safely test and how quickly it can scale. It is a reminder that the project is being pursued by a small group with limited infrastructure, not by a large aerospace contractor or national laboratory.

Pares said QED has run into real hardware constraints. Linear amplifiers are expensive, with units costing around $10,000. He said the team has damaged engines and amplifiers in past attempts, including an attempted quad setup that was too tight in the Faraday cage and resulted in arcing. Those losses matter when a research program is being run on limited funding.

Cable limits are another constraint. Pares discussed coaxial cables rated around 2,000 to 2,500 watts, while saying the aerospace cable he would prefer is rated much higher but would require buying more than the team can afford. He said QED often keeps testing around 1,500 watts because of equipment limits, heat, licensing, and practical caution.

These details make the story warmer because they show the work as a frontier-engineering effort, not just an abstract claim. Pares talks about AI, warp fields, and Bluebird, but he also talks about blown amplifiers, cold garages, hot garages, cable ratings, and volunteer engineers. The VEM Drive’s next chapter depends not only on theory, but on funding, space, hardware, safety, and the ability to keep experiments running.

Direct Drive, Better Electronics, and Future Hardware

One future direction Pares described is moving away from long RF cable runs and toward more direct control of the fractal elements. In the Q&A, he discussed the idea of putting miniature linear amplifiers directly behind each tripole field. That would allow the tripoles to be controlled independently and could eliminate some of the tuning and cable limitations in the present setup.

Pares said QED is aware of this path but that the development cost is significant. He estimated that designing such a miniaturized module could cost around $200,000. For a small company and volunteer-driven project, that is a major step. The idea is clear, but the resources are not yet in place.

This point helps explain why the current work remains incremental. Pares may have a long-range vision of spacecraft, artificial gravity, and rapid planetary travel, but the immediate engineering problems are more concrete: better power modules, better field control, better thermal handling, stronger test rigs, safer high-power setups, and improved measurement tools.

Sherman’s lower-power tests fit into this reality. Pares said Sherman worked with smaller, lower-power, CubeSat-style systems partly because they were safer and less likely to damage expensive equipment. Sherman himself joked that avoiding blown amplifiers was one reason to stay at lower power. That exchange captured the practical tradeoff: lower-power tests are less dramatic, but they are more accessible for replication.

Bluebird and the Larger Vision

The most visual future milestone in Pares’s presentation was Bluebird, a 7-by-7-foot prototype craft located at Council Bluffs Airport. Pares said he wants to mount a VEM engine on the craft and lift it off the ground. For QED, Bluebird represents the move from instrument movement and bench-scale testing toward a more visible demonstrator.

Bluebird should be presented as a demonstrator, not a finished spacecraft. Its purpose is to test whether the effects Pares reports in smaller systems can be scaled into a platform that visibly lifts. Whether the attempt succeeds or leads to another round of redesign, it gives the VEM Drive program a tangible next objective.

Pares’s longer-range goals are much larger. He spoke about satellite station keeping, retrofitting satellites, moving material to space, launching without booster rockets, traveling quickly to Mars, creating artificial gravity, and eventually faster-than-light travel. Those claims should be framed as Pares’s vision rather than established capability.

Sherman’s path is more immediate and measurement-focused. His next steps are a larger rig, stronger PCB fractal arrays, and multipoint interferometry. That contrast is useful. Pares is trying to scale the dream toward Bluebird and beyond. Sherman is trying to make the basic effect easier to reproduce and measure.

The APEC Discussion: Open Questions and Next Tests

The Q&A after Pares’s presentation was collaborative and wide-ranging. Participants asked about field overlap, exponential versus linear increases, adjustable apertures, three-dimensional fractals, material optimization, RF polarization, pulsed operation, cable ratings, direct-drive electronics, clock tests, Faraday rotation, and the relationship between redshift and force.

The theory discussion was also open-ended. Some participants noted that Pares is describing the VEM Drive phenomenologically: he is reporting what he sees, building parameters around the observations, and trying to model the behavior. They also noted that a complete theoretical structure in conventional terms — such as solved field equations or a stress-energy tensor explanation — is not yet in place.

That exchange should be treated as constructive, not dismissive. In frontier research, especially in a community like APEC, unusual claims often begin with apparatus and observation before they are translated into more formal theoretical language. Pares’s own emphasis remains empirical. He is focused on what the engines appear to do, how they respond to geometry, and how the behavior can be reproduced.

Several measurement ideas emerged as possible next steps. Participants discussed correlating force and redshift during latency, using rubidium clocks to look for tiny timing changes, testing Faraday rotation, improving interferometry, optimizing materials, and probing field overlap regions more precisely. These are exactly the kinds of tests that could make the next VEM Drive update more useful to outside observers.

Between Mainstream Physics and Open-Minded Reporting

A responsible story about the VEM Drive has to acknowledge that the interpretation remains outside mainstream propulsion science. Conventional physics does not currently treat a 48-volt, RF-driven fractal device as a demonstrated method for compressing spacetime or producing propellantless thrust. That does not mean the story has to be hostile. It means the language has to be careful.

The right tone is open but precise. Pares says the VEM Drive compresses the fabric of space. Sherman says he reproduced repeatable directional movement. QED presents videos as documenting experimental effects. The broader scientific community has not accepted the system as a validated propulsion technology. All of those statements can sit together in the same story.

This is why the phrase “claims” is useful but should not be overused coldly. Pares and Sherman are not merely speculating in the abstract; they are describing hardware, videos, software, test rigs, power levels, geometry, and next experiments. At the same time, extraordinary propulsion claims require careful measurement before they can become engineering facts.

The most accurate frame may be this: the VEM Drive is not yet a settled breakthrough, but it is an active experimental program whose developers believe they have observed repeatable effects worth documenting and refining. That frame gives readers room to be curious without asking them to accept more than the evidence currently establishes.

An Unfinished Experiment With a Widening Toolkit

By the end of the APEC session, the VEM Drive came across less as a single machine than as a growing research ecosystem. There are fractal engines, RF systems, field maps, AI tools, startup software, videos, scale tests, planned interferometers, direct-drive electronics concepts, proposed clock tests, and a prototype craft waiting for a lift attempt.

Pares’s AI modeling gives that ecosystem a new center of gravity. The software allows him to visualize configurations, explore fractal iterations, probe field overlap, and imagine multi-engine arrays with greater precision. For a project built through years of empirical testing, that is a notable change.

Sherman’s replication effort gives the story a second center of gravity. His lower-power test rig, 10 successful tests in a row, and planned measurement upgrades make the VEM Drive story less dependent on one lab’s demonstrations. His work does not settle the question, but it helps move the project toward the repeatability that any serious experimental claim needs.

For now, the VEM Drive remains a question with software, fractals, videos, field maps, and replication rigs attached. Pares and Sherman believe that question is pointing toward a new form of propulsion. The next chapter will depend on how clearly they, QED, and any outside experimenters can measure what they believe they are seeing.

References

The VEM Drive: AI Modeling | David Pares

The VEM Drive: Replication Experiments | Reid Sherman

Quantum Electro Dynamics R&D

Quantum Electro Dynamics R&D — VEM Drive

Quantum Electro Dynamics R&D — Videos

Quantum Electro Dynamics R&D — Interferometer Experiments

Quantum Electro Dynamics R&D — Cavendish Experiments