Voice Commands for Hands-Free Scanning

If you have ever tried to use a phone app while holding a flashlight in one hand and bracing a toolbox lid open with the other, you understand why hands-free operation matters. AR Toolbox includes a voice command system designed specifically for field conditions where touching the screen is inconvenient, impractical, or unsafe. With a short set of spoken commands, you can trigger scans, save results, navigate between containers, and control the core workflow without ever putting down what you are holding.

Core Voice Commands

The voice command system is built around a small, memorable set of triggers that map to the most common actions in the scanning workflow. You do not need to memorize a long list of phrases or use specific wording. The system is designed to recognize natural variations of each command.

Saying "scan" triggers a detection pass on the current camera frame. The app processes what the camera sees, runs the YOLO model, and produces results exactly as if you had tapped the scan button on screen. You will hear a brief audio confirmation when the scan is initiated and another when results are ready. This audible feedback is essential because you may not be looking at the screen when you issue the command.

Saying "save" commits the current scan results to your inventory under the active container. If you have not reviewed the results yet, the app saves all detections above the confidence threshold. If you have already reviewed and made corrections on a previous pass, those corrections are preserved. The save confirmation includes a spoken summary, such as "Saved 12 items to Red Toolbox," so you know exactly what happened without checking the screen.

Saying "next container" advances to the next container in your list. This is particularly useful when you are scanning multiple storage units in sequence, like the compartments of a service truck. Instead of tapping through menus to switch containers, you simply speak the command and the app moves on. A spoken confirmation tells you the name of the newly active container so you can verify you are in the right place.

Additional commands include "review" to open the review screen for the most recent scan, "undo" to roll back the last save action, and "status" to hear a spoken summary of the current container's inventory state, including any missing expected items.

Voice Feedback and Noisy Environments

Every voice command produces audio feedback confirming what the app understood and what action it took. This two-way communication is what makes hands-free operation genuinely practical rather than a gimmick. You speak a command, the app confirms it heard you correctly, performs the action, and reports the outcome. If the app does not understand a command, it says so, prompting you to repeat rather than silently doing nothing or taking the wrong action.

Field environments are rarely quiet. Workshop compressors, truck engines, power tools running nearby, and general job site noise all create challenging conditions for voice recognition. AR Toolbox addresses this with a dedicated noise-filtering layer that isolates speech from background sound. The voice recognition engine is tuned for the specific command vocabulary rather than general-purpose dictation, which significantly improves accuracy in noisy settings because it only needs to distinguish between a handful of short phrases rather than transcribing arbitrary speech.

You can also adjust the microphone sensitivity in settings. If you find that background noise is triggering false commands, increasing the sensitivity threshold means you need to speak more directly and clearly, which reduces false positives. Conversely, in a quiet shop environment, you can lower the threshold so the app picks up commands spoken at a normal conversational volume from a few feet away.

Voice commands are not a secondary feature bolted on as an afterthought. They were designed alongside the core scanning workflow because the people who need this app most, technicians in the field, are the same people who are least likely to have a free hand.

Practical Scenarios for Voice Scanning

Consider an HVAC technician inventorying a rooftop unit's tool compartment. They have climbed a ladder with their phone propped in a pocket or clipped to a lanyard. They pull open a compartment, hold their phone so the camera can see inside, and say "scan." The app identifies the tools, confirms the count, and the technician says "save" and moves to the next compartment with "next container." The entire process for multiple compartments takes a few minutes with zero screen interaction.

Or picture an electrician wearing insulated gloves at the start of a shift. Taking gloves off to tap a phone screen is a minor annoyance that adds up over the course of a day. With voice commands, the gloves stay on. The electrician opens their tool bag, says "scan," reviews the spoken summary, confirms everything is present, and gets to work. At the end of the job, a quick "scan" and "status" check confirms nothing was left behind.

Even in a shop environment, voice commands speed up the workflow. When you are sorting tools back into their containers after a project, you can keep both hands on the tools while your phone, propped on a stand or leaning against a surface, handles the scanning and saving through voice commands alone. It turns a tedious reorganization task into something that flows naturally with the physical work you are already doing.

What's Next

Voice commands make scanning faster and more accessible, but the real organizational power of AR Toolbox lives in its container management system. In the next post, we will dive into how containers work, from creating and labeling them to nesting containers within containers, generating QR codes, and tracking container history over time.