#design technology

i miss doing that

Some context below the cut if you’re interested (cus I am)

Journal: Coding with Lego Mindstorm

Last February I participated in a Lego Mindstorm robotics workshop, over the winter I assisted with a STEMFest robotics challenge at a local high school, and on Wednesday 7 December I participated in a second Lego Mindstorm robotics workshop.  In the 7 December workshop there were students present who had no prior experience with Lego Mindstorm and a few students like me who had attended the previous workshop in February.  Therefore the workshop was designed to cater to two different levels of experience.

Although I had participated in the prior workshop I still found the recent workshop quite challenging. Perhaps this was partially due to the amount of time that elapsed between workshops.  However, I found that the workshop was very well structured to support the different learning levels.  Providing those types of extension activities is definitely something I will keep in mind for future lesson planning.  The basic instruction was also different from the previous workshop though and so provided its own challenges.  Overall I found it to be an enjoyable experience in problem solving.

As for the value of Lego Mindstorm for teaching robotics, as discussed in the workshop, that is a little more questionable.  The main value is in making coding seem fun and accessible.  There are of course other many other ways of doing this, such as Scratch, Hopscotch, Swift Playgrounds and a variety of other basic coding apps that allow users to create their own games.  Lego Mindstorm, however, has the added benefit of connecting to popular children's toys.  It is also more tactile and the effects are observable in a real world sense, rather than simply on a screen.  When I assisted with the STEMFest robotics challenge there were students who generally had limited interest in computers and programming walking by and expressing jealousy and curiosity about what their schoolmates were doing with the Lego Mindstorm robots. 

Capturing students' interest, then, would seem to be the primary benefit of the Lego Mindstorm robots. However, that benefit may not necessarily justify the cost of the robots, which is quite high.  The coding language is also fairly limited and while many of the functionalities and basic logic systems behind it carry over to other languages, the language itself has no applicability beyond the Lego Mindstorm environment.  It is a tool for teaching concepts or an environment to allow students to experiment within. It is not, however, teaching a versatile programming language, nor does it give students a product that they can use or develop on their own computers at home, as they might with a program or game developed in one of the free apps like Scratch.

There are definitely both pros and cons to the Lego Minstorm package.  I had a lot of fun using it and so did the students at the high school I was assisting at over the winter.  If I happened to have access to the robots and a classroom full of computers at a future school I would certainly be happy to use it in some of my classes.  On balance, however, it may not be worth the cost of acquisition given the alternatives that are available.

References

Architectural firms face mounting pressure to deliver projects faster, meet stringent sustainability requirements, and manage increasingly complex client demands—all while maintaining design excellence. Artificial intelligence offers solutions to these challenges, but successful implementation requires strategic planning and alignment with existing workflows. Firms that approach AI adoption methodically, focusing on specific pain points rather than wholesale transformation, achieve better outcomes and faster return on investment.

The first step in leveraging AI in Architectural Practice is identifying high-impact use cases within current operations. Rather than attempting to automate entire processes, firms should target repetitive, time-consuming tasks that drain resources from higher-value design work. Common starting points include automating code compliance checks during design documentation, using AI to generate preliminary cost estimates during conceptual design development, and applying machine learning to optimize HVAC and lighting systems for LEED certification. These applications deliver measurable efficiency gains without disrupting core design workflows.

Establishing Data Foundations for AI Success

AI systems are only as effective as the data they are trained on, making data quality a prerequisite for successful implementation. Architectural firms generate vast amounts of project data—BIM models, construction drawings, RFI logs, post-occupancy evaluation reports—but this information is often stored inconsistently across platforms and projects. Before deploying AI tools, firms must standardize data formats, centralize project repositories, and establish protocols for capturing information throughout the project lifecycle.

Building information modeling platforms provide a natural foundation for this effort. By ensuring that all team members follow consistent BIM protocols—naming conventions, model organization, parameter definitions—firms create structured datasets that AI algorithms can analyze effectively. Practices like Foster + Partners and Skidmore, Owings & Merrill have invested heavily in these data infrastructure improvements, recognizing that clean, organized project data enables not only AI applications but also better project delivery outcomes overall.

Integrating AI into Design and Delivery Workflows

Once data foundations are in place, firms can begin integrating AI tools into specific workflow stages. During site analysis, AI can process geospatial data, environmental conditions, and zoning regulations to identify optimal building orientations and massing strategies. In conceptual design development, generative design platforms can explore thousands of layout options based on program requirements, circulation patterns, and structural efficiency. These tools do not replace architectural judgment—they expand the solution space architects can consider within tight project timelines.

For firms working on design-build or design-bid-build contracts, AI can streamline the transition from design to construction. Machine learning models trained on historical project data can predict constructability issues before they arise, flagging design elements likely to generate RFIs or require value engineering. During construction oversight, AI-powered image recognition can monitor site progress by analyzing drone footage or construction photos, comparing actual conditions against BIM models to identify discrepancies early. Implementing these capabilities often requires partnering with custom AI solutions tailored to a firm's specific project types and delivery methods.

Training Teams and Managing Change

Technology adoption fails when teams lack the training and support needed to use new tools effectively. Architectural staff, particularly those with years of experience in traditional design methods, may be skeptical of AI-driven recommendations or uncomfortable relying on algorithmic outputs. Firms must invest in comprehensive training programs that explain not only how to use AI tools but also the logic behind their recommendations. Transparency builds trust—when architects understand how an AI model arrived at a design suggestion, they are more likely to engage with the technology constructively.

Change management should also address workflow integration. AI tools should enhance existing processes rather than requiring architects to adopt entirely new working methods. For example, if a firm uses Revit for design documentation, AI tools should operate as Revit plugins or integrate seamlessly through APIs, allowing architects to access AI capabilities without switching between platforms. Kohn Pedersen Fox Associates and similar firms have found that embedded integrations reduce friction and accelerate adoption across project teams.

Measuring Impact and Iterating

Successful AI implementation requires ongoing measurement and refinement. Firms should establish clear metrics before deploying AI tools—time saved in design documentation, reduction in RFIs during construction, improvements in energy performance modeling accuracy—and track these metrics consistently. Regular reviews allow teams to identify which AI applications deliver value and which require adjustment or replacement. This iterative approach ensures that AI investments align with business objectives and project delivery goals.

Conclusion