As artificial intelligence moves from experimental classroom tool to a subject students are expected to understand, educators face a more difficult question than whether learners should use AI: How should they learn to build, question and evaluate it? At TAMYZ FORUM 2026 in Astana, Kazakhstan, DFRobot highlighted an education model built around open-source hardware, AI experimentation and robotics, arguing that practical projects can help students move from consuming AI tools to understanding how data, models and algorithms shape their results.
The rapid adoption of generative AI has created an awkward gap in education.
Students increasingly encounter AI through chatbots, search engines, image generators and automated learning tools. Yet understanding how those systems work—and knowing when their outputs should be trusted—is a very different skill.
That distinction was central to DFRobot’s presentation at TAMYZ FORUM 2026, an education forum held August 17–18 in Astana, Kazakhstan.
Sandy Zhang, vice president of DFRobot, delivered a keynote titled “Advancing AI Education Through Practical Innovation,” outlining the company’s approach to AI education through open-source hardware, robotics and hands-on experimentation.
The forum, organized by the Astana City Education Department and Astana City Methodological Center, focused on the intersection of education policy, artificial intelligence and human capital. DFRobot said the event attracted around 3,000 in-person and 20,000 online participants across 15 thematic sessions.
The company’s central argument is straightforward: AI education should progress from teaching students how to use AI tools toward teaching them how AI systems work.
That means exposing learners to data collection, model training, testing and deployment rather than presenting artificial intelligence as an opaque service that simply produces an answer.
The approach reflects a broader shift in AI education.
Governments and school systems around the world are developing AI literacy frameworks, but their priorities differ. Some focus heavily on technical competencies, while others emphasize ethics, governance, digital citizenship and responsible use.
The common challenge is that schools need to prepare students for technologies that are changing faster than traditional curriculum cycles.
For DFRobot, hands-on hardware offers one way to make the underlying concepts easier to teach.
The company highlighted its HUSKYLENS 2 AI Vision Sensor, which can be used in classroom projects involving data collection, labeling, model training and real-time visual recognition.
A project could, for example, have students gather images, label different objects and train a model to recognize them. Different student groups can use different datasets and subsequently obtain different results.
That variation creates a useful teaching moment.
Instead of simply telling students that training data influences an AI model, educators can demonstrate the relationship experimentally. Students can observe errors, change the data, retrain the model and compare the results.
In other words, AI becomes something they can inspect and modify rather than simply query.
That distinction is increasingly important as AI-generated information becomes embedded in everyday decision-making.
A student who understands that model output depends on training data, labeling and system design is better positioned to question an AI response than someone who has learned only how to formulate prompts.
The same projects can introduce more complicated topics, including data privacy, algorithmic bias and responsible AI deployment.
This approach places DFRobot in a broader ecosystem of AI education providers, robotics companies and developer platforms. Microsoft’s educational AI initiatives, Google’s AI learning resources and NVIDIA’s academic programs similarly reflect growing demand for AI skills, although their approaches range from software education to advanced computing and research.
DFRobot’s differentiation lies heavily in physical computing.
Its ecosystem combines robotics, sensors, AI hardware and development modules with software and educational resources. The company’s Mind+ programming platform is designed to lower the barrier to programming and hardware development, while curriculum materials, teacher training and project resources are intended to help schools translate technology into classroom activities.
That integrated model matters because hardware alone rarely solves an education problem.
A school can purchase an AI sensor, but teachers still need lesson plans, training and a practical way to connect the device to learning objectives.
The challenge becomes particularly significant in regions where educators may be expected to introduce AI without having extensive technical backgrounds themselves.
DFRobot says its ecosystem is designed to serve K–12 students, teachers, universities, developers, makers and researchers. That creates a pathway from introductory STEM projects toward more advanced experimentation.
The company’s international partnerships are part of that strategy.
DFRobot has participated in initiatives involving UNESCO and the Arab League Educational, Cultural and Scientific Organization (ALECSO), while working with institutions including Kathmandu University in Nepal and Chiang Mai University in Thailand on curriculum development and teacher training.
The emphasis on international collaboration is significant because AI education is becoming a workforce-development issue as much as an education issue.
Countries are competing for AI talent while simultaneously trying to ensure that students understand the technology’s social and ethical implications. That creates demand for education systems capable of developing both technical competence and critical thinking.
There is also an important economic dimension.
Teaching students to use AI applications may prepare them for today’s workplace. Teaching them to understand data, test models and build AI-enabled systems could prepare them for technologies that have not yet been developed.
That is the rationale behind the “developer mindset” DFRobot described at the forum.
It does not necessarily mean every student needs to become a machine-learning engineer. Rather, students should gain enough technical understanding to experiment, identify limitations and turn technology into practical solutions.
For enterprise technology leaders, the trend has implications beyond schools.
Companies adopting AI increasingly need employees who can evaluate AI outputs, understand basic model limitations and work effectively with automated systems. That makes AI literacy part of the broader enterprise skills pipeline.
The education technology market is consequently moving toward a more integrated model in which AI software, physical computing, curriculum and teacher development converge.
DFRobot’s strategy is one example.
Its larger proposition is that lowering the technical barrier to experimentation can broaden participation in AI development—from professional engineers to students, teachers and makers.
Whether that translates into sustained learning outcomes will depend on factors that hardware alone cannot solve, including teacher preparation, curriculum quality, access to computing resources and how schools measure AI competencies.
But the direction is clear.
As AI becomes ubiquitous, education systems will have to decide whether students should primarily become better users of AI—or become informed participants capable of understanding, challenging and creating with it.
DFRobot is betting on the latter.
Market Landscape
The AI education market is evolving beyond standalone coding courses and generative AI literacy programs.
Four areas are increasingly converging:
AI literacy: Helping students understand AI concepts, limitations, ethics and responsible use.
AI development: Teaching programming, model training, data handling and experimentation.
Physical AI and robotics: Connecting machine-learning concepts with sensors, robots and real-world environments.
Teacher enablement: Giving educators curriculum, training and tools needed to introduce AI effectively.
This shift mirrors the broader AI economy. Microsoft, Google, NVIDIA, Amazon and universities are investing in AI education and workforce development, while robotics and open-source hardware companies are creating more accessible development environments.
For schools and governments, the strategic challenge is not simply purchasing AI tools. It is creating a sustainable education infrastructure around them.
That includes teacher training, curriculum standards, privacy policies and opportunities for students to build practical projects.
The strongest AI education models may therefore look less like traditional software training and more like hands-on engineering education, combining computational thinking with real-world experimentation.
Top Insights
- DFRobot is promoting hands-on AI education through open-source hardware, robotics and model-training projects designed to help students understand AI beyond everyday tool usage.
- HUSKYLENS 2 lets students collect data, train models and test visual recognition, making concepts such as bias and model performance tangible through experimentation.
- The company’s integrated ecosystem combines hardware, Mind+ software, curriculum, teacher training and competitions, targeting schools, universities, developers and makers.
- International education partnerships illustrate the growing role of AI literacy and practical technology skills in national workforce-development strategies.
- The broader education market is shifting from AI tool familiarity toward technical understanding, critical evaluation and responsible creation with artificial intelligence.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
