How To Ask Robot Building Questions

The answer is only as good as the question

When you ask a well-structured question on the Synthiam forum, you help both yourself and everyone who searches for the same issue later. Clear questions lead to faster, more accurate replies from the community and better results from Athena, our AI support agent.

Athena is powered by an LLM (Large Language Model). Like all LLMs, it depends heavily on your input. If the goal is unclear or key details are missing, the answer quality drops quickly—garbage in, garbage out. The more precisely you describe the outcome you want, the easier it is for Athena (and humans) to focus on the right solution.

What Athena answers best

Athena performs best when your question has a single, clearly defined outcome and enough supporting details to reach that outcome. If your post includes multiple topics, multiple goals, or jumps between unrelated problems, Athena will tend to “gravitate” toward one theme and may miss what you actually want. Humans can ask follow-up questions and reframe the problem; Athena can too, but it is still limited by the clarity and structure of the thread.

• Best: One topic, one goal, clear success criteria, and the details needed to reproduce or understand the situation.
• Harder: Multiple topics in one thread, unclear goal, or a list of outcomes (“also while you’re here…”).
• Worst: A long thread where the subject changes mid-stream and the expected outcome is never clearly stated.

Always include a clear outcome

Before you post, define the outcome you want in one sentence. This is the “finish line” for your question—what success looks like. Include every detail that affects that outcome: hardware, versions, steps, expected behavior, and actual behavior.

Keep one thread = one topic

If the subject changes during a question thread, start a new thread. This is important for two reasons:

• Clarity for people: It keeps the discussion searchable and prevents solutions from getting buried in unrelated replies.
• Clarity for Athena: LLMs build context from the thread and tend to anchor to the dominant topic. When a thread contains multiple topics, the model has to guess which one matters most. When a thread contains multiple outcomes, it has even more difficulty focusing on the correct finish line.

If your new question is not 100% the same problem and the same desired outcome, make a new post. Link to the old thread if helpful.

Guidelines for asking a question

• Start with the outcome:
  • Write one sentence describing what you want to happen (success criteria).
  • Write one sentence describing what is happening instead (actual behavior).
• If it’s an issue or bug:
  • Provide the exact error message (copy/paste).
  • List steps to reproduce it (numbered steps are best).
  • Include environment details: OS, ARC/ARCx version, firmware version, device model, connection type, and any connected hardware.
  • Include screenshots/logs if they show the problem clearly.
• If it’s “how do I do this?”:
  • Explain what you are trying to build and why (one paragraph).
  • List what you’ve tried and what result you got.
  • Mention what documentation you already reviewed, like the Getting Started Guide.
• If it’s a programming question:
  • Provide complete code that reproduces the issue (not partial snippets).
  • Clearly state the exact outcome and behavior you want from the code.
  • Explain what you did that triggered the issue and what you expected to happen.
  • Include any inputs, sample data, and the exact output you got.
• Make it easy to scan:
  • Use short paragraphs, headings, and bullet points.
  • Use code blocks for code and error logs.
• Don’t hijack threads: Ask a new question rather than appending to an existing thread unless your question is 100% related and has the same outcome.

Why properly asking a question matters

The Synthiam forum works because questions and answers remain easy to understand and easy to find. When posts are clear, the community can respond quickly, Athena can provide better guidance, and future users can discover the solution through search.

Every well-asked question becomes a reusable resource. When questions are vague, mix topics, or skip key details, answers become guesses. Over time that turns the forum into a messy pile of half-solutions where nobody can reliably find what they need.

Clear questions are how we keep the forum fast, searchable, and genuinely helpful.

How Athena can help

Athena is designed to answer new, well-documented questions—especially those with a clear outcome and enough context to make the problem reproducible. If your post includes the outcome, environment, steps, and supporting details, Athena can usually give a strong first answer and suggest the best next diagnostic steps.

If you change subjects mid-thread, create a new thread. One thread per topic helps Athena stay focused and helps everyone else find the right answer later.

Sample questions

Scenario 1: Troubleshooting an issue

Outcome: “I want ARC to connect to my EZ-B v4 over Wi-Fi without timeouts.”

Question: “I’m using Synthiam ARC vX.X.X with an EZ-B v4. When I connect, I get: ‘Connection failed: Timeout.’ Steps to reproduce: (1) Power EZ-B, (2) connect PC to EZ-B Wi-Fi, (3) press Connect in ARC. Expected: ARC connects within 5 seconds. Actual: timeout after X seconds. OS: Windows 11. Firmware: vX. Router/AP: (if applicable). I’ve restarted the EZ-B and PC and tried a second device.”

Scenario 2: Learning how to do something

Outcome: “I want my robot to wave once when motion is detected by the camera.”

Question: “I want my robot to wave when the camera detects motion. I have the camera connected and visible in ARC. I followed the camera tutorial and can see the feed, but I’m unsure how to trigger a wave script reliably. Can someone show a simple approach (or skill combination) that detects motion and runs a wave script once per motion event?”

Scenario 3: Seeking design advice

Outcome: “I want to choose servos that can lift a specific payload with smooth motion.”

Question: “I’m designing a robot arm and want it to mimic human motion as closely as possible. The payload at the end effector is ~X grams, arm length is ~Y cm, and I want smooth movement (not speed-focused). What servo torque range should I target, and are there any prebuilt arms or example projects that integrate well with Synthiam ARC?”

Additional tips

• Search the forum before posting—your question may already be answered.
• Read the Getting Started Guide because it addresses most common questions.
• Use proper grammar and formatting so your post is easy to read and easy to answer.