Building SandWorm, one test at a time.

It can be seen that it is the lid of an aluminum or iron can.

The new generation of chassis solves the obstacle-crossing issues we previously encountered, enabling us to test in more places and diverse environments more effectively.

It has improved our obstacle-crossing performance, and we will test it immediately after installation.

Having solved the problem of metal detection interference, we turned our attention to a new problem that emerged in low-speed tests: the rover struggled to overcome obstacles when moving slowly. This journal records our findings and remediation efforts.

Observed problems

In controlled tests on uneven terrain, the Gen2 prototype performed well at moderate speeds, but struggled slowly over rocks, roots and small slopes. At low speeds, the rover sometimes loses power and is unable to climb obstacles that are supposed to be within its capabilities. This is a concern because precise, slow movement is essential for a full scan.

The investigation found

We investigated two possible reasons: power output and control algorithms. In terms of power, we found that the motor's output torque at low RPM was not enough to maintain traction on rolling ground. On the algorithmic side, the speed control ring is too conservative at low speeds, causing a delay in accelerator response when the rover encounters resistance.

Measures for improvement

We made adjustments in two areas. First, the motor drive parameters were recalibrated to provide a more consistent torque output in the full speed range, especially at low speeds. Second, the speed control algorithm was updated to respond more quickly to sudden load changes, and a specific low-speed compensation curve was added to prevent deceleration on slopes or obstacles.

Result

Now rovers are more reliable when dealing with obstacles at low speeds. In repeated tests on mixed terrain, it was able to climb small rocks and uneven ground without losing speed or straying from its established path. This improvement is particularly pronounced when slow scans pass, where stability and accuracy are the things that are most valued.

Follow-up plans

With this issue resolved, the Gen2 prototype is ready for its first full outdoor field test. We will evaluate the full system - including mobility, detection and data recording capabilities - to operate in conditions as close to a real treasure hunt as possible. When the test is complete, we will share the results here.

Through our continuous testing, we dug up a lot of different iron things on the lawn.

Over the past few weeks, we have focused on resolving the issue of interference between the metal detection system and the chassis of the detection vehicle. We are pleased to report that this problem has now been resolved.

Modified content.

After conducting a series of tests, we made two key adjustments. First, we rearranged the detection coils to keep them away from the motor and high-current wiring of the detection vehicle. Second, we modified the structure of the chassis section to reduce the presence of metal components near the sensing area. Additionally, we added shielding to several electronic components to reduce electromagnetic noise.

Result

Now, even when the detector vehicle is moving under normal operating conditions, the detection system can produce clean and stable readings. Background noise has been significantly reduced, and the system can clearly identify buried metal objects during controlled tests.

the next step

The interference issues are a thing of the past, and the Gen 2 prototype is ready for its first real outdoor field test. We will take it to a mixed terrain site to evaluate the entire system under conditions as close to a real treasure hunt scenario as possible—including mobility, detection, and data recording capabilities. We will release the results after the test is completed.

Based on the issues identified during our initial driving tests, we redesigned the steering system. We are pleased to report that the steering response of the Gen 2 detector has significantly improved. Following the completion of these improvements, we took the next step: installing the metal detection system onto the platform for the first time.

Operation Content

We tested the updated steering configuration on the same mixed terrain as before, and the exploration vehicle was able to navigate more narrow paths with improved maneuverability. Once the handling performance met the required standards, we installed the metal detection module and conducted a series of static and low-speed tests.

Issues identified.

Although both systems functioned normally when operated independently, we immediately noticed interference when they were running simultaneously. The metal detection readings became noisy and unstable, particularly when the vehicle’s motor was in operation or when the detection coil was close to chassis components. It is clear that the vehicle’s structure and electronic components introduced unnecessary noise into the detection signals.

Future plans.

It is necessary to isolate the detection system from the electrical and structural interference of the detection vehicle. We will address this challenge by addressing issues such as coil layout, shielding measures, and adjustments to the chassis design. This is a critical hurdle that must be overcome before meaningful field testing can begin.

We recently completed the first mobility performance test of the second-generation SandWorm exploration vehicle. The objective was straightforward: to take a prototype equipped with a new wheel-based chassis onto real terrain and assess its handling characteristics before installing any exploration hardware.

Test content.

The Gen 2 platform features a redesigned wheel-driven system designed to enhance mobility on mixed terrains. We brought it to a nearby testing area that included grass, compacted soil, and various uneven sections to assess basic movement, stability, and steering performance.

Issues identified.

The exploration vehicle performed well while driving in a straight line, able to navigate moderately bumpy terrain smoothly. However, we soon discovered limitations in the steering system – the turning radius was much larger than expected, and it was difficult to smoothly make sharp changes in direction. In confined spaces, the vehicle was unable to effectively adjust its position, which could pose problems in actual searches.

Future plans.

The primary task currently is to modify the steering geometry and control logic to enable the detection vehicle to achieve smaller and more predictable turning radii. Once this issue is resolved, we will move on to the integration phase of the metal detection system. Further developments will be continuously updated in the log.