AI Assist: Voice and Text Diagnostics

AI Assist: Voice and Text Diagnostics

What You'll Learn

  • How to use voice and text interaction to get AI-powered diagnostic guidance from live measurement data
  • How AI Assist differs from mQ Assist (I4) and why that distinction matters
  • What the output categories are and how to read them
  • How to use the homeowner explanation to communicate findings to customers
  • What AI Assist will and will not answer

What You'll Need

  • Subscription: Premier Services (AI Assist is a Premier feature)
  • App version: measureQuick v3.6 or later
  • Device: iPhone (iOS 15+) or Android phone/tablet (Android 10+) with microphone access enabled
  • Probes: Bluetooth smart tools connected and streaming live measurement data
  • Time: 10 minutes to read; used continuously during diagnostics

Step-by-Step Guide

Step 1: How AI Assist Differs from mQ Assist

If you have used mQ Assist (covered in mQ Assist AI Diagnostics), you already understand the concept of getting diagnostic guidance from measureQuick. AI Assist in version 3.6 is a different system built on a different foundation.

mQ Assist (I4) uses pattern matching on diagnostic flags. It looks at which pass/fail subsystems are flagged and suggests what the combination of flags might mean. It works from the binary results that measureQuick has already computed.

AI Assist uses a proprietary AI engine that analyzes live measurement data directly. It does not wait for pass/fail flags. It reads the actual temperature, pressure, airflow, and electrical values streaming from your probes and interprets them in real time. As Jim Bergmann describes it: "We built our own AI engine in measureQuick, so it runs off of measurements instead of you inputting measurements."

The practical difference: mQ Assist tells you what the diagnostic flags mean together. AI Assist tells you what your measurements mean, what to do about them, and how to explain the situation to a homeowner. AI Assist also accepts follow-up questions by voice or text, making it conversational rather than static.

Both features can be active at the same time. They complement each other.

Feature mQ Assist (I4) AI Assist (I14)
How it works Pattern matching on pass/fail diagnostic flags Proprietary AI engine analyzing raw measurement data
Input Computed diagnostic results Live probe measurements (temperatures, pressures, airflow, electrical)
Interaction Automatic display of guidance text Voice or text conversation
Output Contextual diagnostic suggestion Diagnostic analysis with actionable steps; up to four categories when full data is available
Availability All tiers Premier Services only
Version 3.5+ 3.6+

Step 2: Accessing AI Assist

AI Assist is available from two locations depending on your interface mode.

In mQ+ (Premier Services interface):

  1. During an active diagnostic session, swipe up or tap the Navigation Drawer at the bottom of the screen.
  2. Tap the Assist button (or Mic button) in the drawer.

📷 mQ+ Navigation Drawer open at the bottom of the screen, with the Mic button highlighted among flashlight, actions, camera, and toolbox options

In mQ Classic:

  1. Navigate to the diagnostics screen during an active session.
  2. Tap the AI Assist button on the diagnostics toolbar.

mQ Classic diagnostics screen with the AI Assist button highlighted in the toolbar

mQ Classic diagnostics screen with the AI Assist button highlighted in the toolbar

In both cases, AI Assist opens a conversation panel with a greeting ("Hi! I'm mQ Assist. How can I help you today?") and three category buttons: Diagnostics, How-To, and Knowledge. You can tap a category to browse pre-built questions, or speak/type your own question directly.

Step 3: Voice Interaction

Voice is the primary way to use AI Assist in the field, especially when your hands are occupied with tools or equipment.

  1. Tap the Mic icon (right side of the text input field) to start listening.
  2. Speak your question clearly. Examples:
    • "Tell me about the system operation."
    • "Why is my superheat so low?"
    • "What should I check next?"
    • "Is this system overcharged?"
  3. AI Assist processes your question alongside the current measurement data.
  4. The response appears on screen as text (see Step 5 for the output format).

You can ask follow-up questions without starting over. AI Assist maintains context from the current session and current measurement state.

📷 AI Assist listening mode active with the mic icon pulsing, measurement data visible in the background

Tips:

  • Speak naturally. You do not need special commands or keywords.
  • Ask specific questions for more useful answers. "Why is superheat low?" produces better output than "What's wrong?"
  • If the environment is noisy (compressor running, fan noise), hold the phone closer to your mouth or use text input instead.

Step 4: Text Interaction

If voice is not practical, you can type questions directly.

  1. Open AI Assist (via the Assist button, Mic button, or diagnostics toolbar).
  2. Tap the text input field at the bottom of the conversation panel (placeholder text reads "Ask about this system...").
  3. Type your question and submit.
  4. AI Assist responds with diagnostic guidance in the same format as voice interaction.

Text interaction is useful when you are in a quiet environment (like reviewing data in the truck) or when ambient noise makes voice recognition unreliable.

📷 AI Assist text input field with a typed question and the AI response displayed above

Step 5: Understanding the Output Categories

When AI Assist has sufficient measurement data and context, it organizes its response into up to four sections. Shorter queries or questions asked without probes connected may return a concise answer instead of the full breakdown. The four sections are:

1. System Analysis

What the data shows. This is a factual summary of the current measurements and what they indicate about system operation. AI Assist reads the live values from your probes, not just the pass/fail flags, so it can identify specific conditions such as "low evaporator load" or "elevated discharge temperature" even before the app flags a formal failure.

Example: "The system is showing 1.9 degrees of superheat with subcooling at 14 degrees on a TXV system. Return air temperature is 65F, indicating low load on the evaporator."

2. What This Means

Interpretation of the data. This section translates the raw analysis into a diagnostic conclusion. It explains the likely cause or condition in terms a technician can act on.

Example: "Low superheat with a TXV suggests the sensing bulb may not be making good thermal contact with the suction line. The evaporator is overfilled with refrigerant. This can mimic an overcharge condition, but the root cause is likely the TXV response rather than excess refrigerant."

3. Recommended Actions

What to check or do next. This section provides specific troubleshooting steps based on the analysis. The steps are ordered by likelihood and ease of verification.

Example: "Check TXV sensing bulb mounting on the suction line. Verify the bulb has proper thermal contact and insulation. Check for airflow restrictions at the filter and return. Verify the equipment profile is correct (tonnage, refrigerant type, metering device)."

4. Homeowner Explanation

A plain-language version of the findings written for a customer who does not have HVAC technical knowledge. This section is designed to be read aloud or shared directly with the homeowner.

Example: "Your air conditioning system has a component that controls how much refrigerant flows through the cooling coil. That component appears to need adjustment. Right now, too much refrigerant is flowing through the coil, which reduces the system's ability to cool and dehumidify efficiently. We need to inspect the sensor that controls this flow and make sure it is properly attached."

Full AI Assist response showing all four sections - System Analysis, What This Means, Recommended Actions, and Homeowner Explanation - with a low-superheat TXV scenario

Full AI Assist response showing all four sections - System Analysis, What This Means, Recommended Actions, and Homeowner Explanation - with a low-superheat TXV scenario

Step 6: Using the Homeowner Explanation

The homeowner explanation section is one of the most practical features of AI Assist for customer-facing communication.

Reading it aloud: When you are standing with the homeowner, you can read the explanation directly from your phone. It uses plain language without jargon.

Building trust: Showing the homeowner the measureQuick screen with the AI analysis demonstrates that your diagnosis is based on measured data, not guesswork. The homeowner sees the measurements, the professional analysis, and the plain-language explanation all in one place.

Justifying repairs: When the recommended action involves a repair, the homeowner explanation provides the "why" in terms the customer can understand. This helps the customer make an informed decision about the repair.

Adding to notes: You can store the AI Assist output in the project's service history. Tap Add to Notes to save the analysis for the current service record. This becomes part of the equipment's service timeline, visible on future visits.

[Visual Reference] The Homeowner Explanation section appears below the technical analysis in the AI Assist results. It uses plain, jargon-free language to describe what was found and what it means for the homeowner. The text is displayed in a distinct card or section, visually separated from the technical summary above. At the bottom of this section, an "Add to Notes" button allows the technician to save the full AI analysis (both technical and homeowner-facing) to the project's service history for future reference.

Tip: If the homeowner asks a question you are not sure how to answer in plain language, type it into AI Assist. For example: "The homeowner wants to know why their system runs all day but the house stays warm." AI Assist will respond with both a technical analysis (for you) and a homeowner-ready explanation.

Step 7: What AI Assist Will and Will Not Answer

AI Assist is restricted to measureQuick and HVAC topics. It is built for diagnostics, not general conversation.

AI Assist will answer questions about:

  • Current system operation based on live measurements
  • Diagnostic interpretation (what readings mean, what to check) - tap the Diagnostics category button to browse topics: Refrigerant, Airflow, Combustion, Electrical, Metering Device
  • Equipment behavior (why superheat is high, what subcooling indicates)
  • Troubleshooting steps for identified conditions
  • How-to guidance for app features - tap the How-To category button
  • General HVAC knowledge - tap the Knowledge category button
  • Homeowner explanations of findings

AI Assist will not answer:

  • Questions unrelated to HVAC or measureQuick (restaurant recommendations, weather forecasts, general trivia)
  • Questions about pricing or quoting repairs
  • Questions requiring information it does not have (customer history, warranty status, equipment age beyond what is in the profile)

If you ask an off-topic question, AI Assist will politely decline and redirect you to HVAC-related topics.


Limitations

AI Assist is a diagnostic aid. It does not replace technician judgment. Keep these limitations in mind:

  • Measurement quality matters. AI Assist analyzes whatever data your probes are sending. If a probe is misplaced, the system has not stabilized, or a clamp is on the wrong line, the AI analysis will be based on incorrect data. Verify your probe placement before relying on AI Assist output.
  • More probes produce better results. AI Assist works with whatever data is available, but its analysis is more specific and accurate when it has a complete picture. Nine or more physical probes (the same threshold required for a Vitals Score) give AI Assist the most data to work with.
  • It cannot see the physical system. AI Assist does not know about visual conditions (corroded coils, damaged ductwork, water stains, unusual installation configurations). Combine AI Assist output with your own visual inspection.
  • It does not store structured data. The AI Assist conversation is not saved as structured diagnostic data in the test record. You can save it to notes manually via the Add to Notes button, but the four-category breakdown is stored as text, not as indexed fields.
  • Network connectivity. AI Assist requires an active data connection to process queries. If you are in a basement or crawlspace with no cell service, AI Assist will not respond until connectivity is restored.

Tips & Common Issues

AI Assist is not responding

Verify you have an active internet connection. AI Assist processes queries through measureQuick's servers and requires connectivity. If you are in a low-signal area, move to a location with better reception or switch to text input (which uses less bandwidth than voice).

Also confirm you have a Premier Services subscription. AI Assist is not available on free or basic paid tiers.

The AI analysis does not match what I see in the field

AI Assist works from measurement data only. If the analysis seems off, check your probe placement first. Common causes of misleading AI output:

  • Temperature clamp on the wrong line (suction vs liquid)
  • System not yet stabilized (pressures and temperatures still changing)
  • Incorrect equipment profile (wrong tonnage, wrong metering device, wrong refrigerant)
  • Missing probes (AI is working with incomplete data)

Fix the underlying data issue and ask your question again. AI Assist recalculates from the current measurements each time.

Can I use AI Assist without probes connected?

AI Assist requires live measurement data to provide diagnostic analysis. Without probes, there is no data to analyze. You can still ask general HVAC questions, but the value of AI Assist comes from its ability to interpret your specific system's measurements in real time.

What is the difference between AI Assist and the AI Diagnostics Summary?

AI Assist is the interactive voice/text feature you access through the mic button or diagnostics toolbar. You ask questions and get answers.

The AI Diagnostics Summary is the automated analysis that appears when you tap Generate Diagnostics in the mQ+ workflow. It produces the same four-category output (analysis, meaning, actions, homeowner explanation) but does so automatically based on the complete measurement set at that moment. You do not ask a question; the system generates a comprehensive summary.

Both use the same underlying AI engine. AI Assist is conversational and on-demand. The AI Diagnostics Summary is automatic and comprehensive.

How is AI Assist different from ChatGPT or other general AI tools?

AI Assist is purpose-built for HVAC diagnostics and runs on a proprietary measureQuick AI engine. The key differences:

  • AI Assist already has your measurement data. You do not need to type in temperatures, pressures, or other values. It reads them directly from your connected probes.
  • AI Assist is trained specifically on HVAC fault patterns and measurement relationships, not general knowledge.
  • AI Assist refuses off-topic queries. It stays focused on the job.
  • AI Assist produces structured diagnostic output designed for field use, not free-form chat responses.

Related Articles

Prerequisites (complete these first):

Follow-up articles (next steps after this one):

Related in other domains:


Need Help?

If you get stuck or this article does not answer your question:

  • Check the Related Articles section above
  • Contact measureQuick support: support@measurequick.com
  • Schedule a training session with the measureQuick training team
    • Related Articles

    • mQ Assist (AI Diagnostics)

      What You'll Learn What mQ Assist is and how it provides context-aware diagnostic guidance How mQ Assist differs from static pass/fail indicators - it analyzes relationships between measurements, not just individual thresholds When mQ Assist activates ...
    • AI System Profiler

      What You'll Learn What the AI System Profiler does and why it was introduced in measureQuick v3.5 (2025) How to photograph an equipment nameplate so the AI can read it What data the AI populates automatically: manufacturer, refrigerant, tonnage, ...
    • Two-Speed Compressor Diagnostics

      What You'll Learn How two-stage compressors operate and why diagnostics differ between stages How to identify which stage is running during a test The correct testing protocol for two-stage systems in measureQuick How superheat, subcooling, and CTOA ...
    • Refrigerant Charge Diagnostics

      What You'll Learn How measureQuick evaluates refrigerant charge based on metering device type What the pass/fail indicator means and how it is calculated What the field data shows: 56.0% of piston systems fail charge evaluation, 45.4% overall How ...
    • Standalone Mode (No-Project Diagnostics)

      What You'll Learn What Standalone Mode is and how it differs from project-based workflows How to start a standalone diagnostic session without creating a project How to profile a system, take measurements, and run diagnostics in Standalone Mode How ...