Setting the Stage

DTS is almost here – in just one week we’ll be getting together – minus the armor, swords, and medieval lighting – and tackling what Artificial Intelligence (AI) and Machine Learning (ML) means to the AEC industry, and how we can best position our firms and ourselves in this evolving landscape.

 

According to Gartner’s 2017 Emerging Technologies Hype Cycle report, ML is just about ready to take a roller coaster-like dive into the ‘trough of disillusionment’. After lots of talk and inflated expectations, there will be a letdown as the realities of early implementation come to light, and the hype gets put in check. So how can we even out this trough in our firms? What can we propose that demonstrates the positive aspects of AI, while keeping the disillusionment at bay?

 

One idea we’ve discussed in the DTS Committee is to focus our conversations on the next several years, while keeping the longer-term view in the background. Using this framework to think about ‘driverless car’ technology, we know it isn’t mature enough yet to eliminate the need to drive, but today it is useful in providing useful feedback and assistance to drivers.

 

In the AEC space, two things that come to my mind are Proving Ground’s LunchBox ML plugin, and WeWork’s desk layout automation research.  The former is a tool that can be adopted now, and the latter describes a proof-of-concept that could be expanded to address more complex design scenarios.

 

So be prepared for an intense, fun, and hopefully satisfying day and a half of conversations. As you’re preparing for next week, keep the following in mind:

 

* What improvements can we begin over the next twelve months

* How will they benefit our practices, our work, and our lives, and

* How can they inform and accommodate the next wave of technological advancement?

 

Image credit: https://www.pinterest.com/pin/499758889889807127/

AI and the future of the AE Industry

We are fast approaching DTS 2018 on August 7th, and the enthusiasm is building! The committee has been busy finalizing discussion topics and the excitement has been mounting with each meeting.  We look forward to thought-provoking and impassioned dialogues on present day digital practice topics and challenges.

This year, a hot topic that is quickly becoming a pervasive presence in our every day lives is AI. True AI has yet to infect the AE industry though there are lots of examples that come close. What aspects of an Architectural practice or Engineering practice would we want to involve artificial thinking?  How would we “teach” it and what challenges must we overcome to benefit the most from (unbiased? Logical?) Digital thinking? If we were able to record every answer to each day’s design or management questions, what format does that take to make it consumable by a “digital brain”? And how then does that information get regurgitated when it’s needed?

 

We talk about Artificial Intelligence as being the next great disseminator of information in tomorrow’s world, but we seem to skip the part on how we gather the information.  We are seeing web browsers recommending products to purchase, smart cars starting to drive themselves, and autocorrect putting words into my messages that I have no idea what they mean.  The design industry is influenced so much by personal experiences and emotions, that trying to translate that factor into a simple on/off algorithm might appear to be next to impossible.

 

But what is impossible today may not be tomorrow. We must figure out the right questions to ask.  The growing connection between people via technology is allowing us to look into everyone’s daily habits, learn their routines, and now we can attempt to connect this data with the design process.  But we still have the issue, “What questions do we want to ask?”

 

We will be discussing these questions along with many others at DTS 2018, August 7th-8th in St. Louis.  If you already know the answer or you have even better questions, join us! There are a few seats open!

DTS 2018: AI, People Management and Challenging Practice Norms

As I sit down to write this blog post, it’s hard to believe that it is the last post I’ll write before DTS 2018 kicks off in St. Louis on August 7th! I’m very excited about our program this year, the committee has spent our last few meetings discussing how we will run the discussions and in if our own internal conversations are any measure this year’s discussions should be particularly fruitful and interesting. We are working hard to build on the energy of last year and push the envelope to establish the gaps facing our Design Practices in terms of how the practice of design must engage the 21st Century, its more than BIM, that much is for sure, it’s about restructuring our practice and thinking hard about where do we need to be prepared to go. 

 

For example, for too long the conversation has been focused on making our practices more efficient, without really challenging practice norms. How can technology help improve the lives of our design staff, efficiency does not necessarily equate to an improvement of the lives and careers of our users. Nor does technology alone force us to reconsider how we capture and disseminate knowledge, yet technologies like machine learning, artificial intelligence and “generative design” (what exactly is that) are going to force us to adapt, if not we quite literally may find ourselves and our staff out of a job. To take it a step further, what if you asked a Principal, and Project Manager and an Intern “what should AI do for you in your job” you’re likely to get some quite varied answers; but all the answers likely presume that person still has a job. On top of that, the answers from each are more likely to indicate negative consequences for the “other” person’s job. How do we help to meld all these different opinions together into a consensus that not only helps our firms be successful, but helps to establish criteria by which technology companies can design and develop solutions towards? 

 

We need to understand these coming technologies and their impact on our businesses so that know how our roles adapt and how the roles of our practitioners adapt. 

 

If all of this sounds interesting please consider joining us at DTS 2018, we still have open slots available and we would love to have you. 

 

Nirvana Workflow: What is it and is it possible?

Recently I came across a company that was a totally vertical entity, developing, designing,  constructing and operating high end residential towers… Wait for that to sink in…  Nearly every aspect of their process was done in house. This is a highly unusual circumstance in our industry, nye unprecedented, and one that is entirely fascinating from a design technology perspective.

 

This brought me down the road of a thought experiment.  What would a digital workflow look like if the only requirements were that it delivered a building and a digital dataset for facilities maintenance at the end?  If there were no technological restraints except that only commercially available tools could be used, these tools being readily available to all parties involved, and that the parties would only need to deliver their core requirements. For example,  Code review would only be required to validate that the design met code, and this could be done with digital tools.  This gets to the idea of what we’ve termed this year as the “nirvana workflow” and is one of the discussions planned for the Summit. What would this workflow look like? Is it even currently possible? Where would custom programming be required to keep the data flowing between development, concepts, model(s), and maintenance? What aspects of current conventional workflows fall away or become entirely unnecessary? What opportunities for automation are there? Where are areas that AI could step in to enhance it further?

 

For folks like us, this thought experiment is exciting and perhaps even liberating. The outcomes of such an exercise may identify opportunities to explore further, perhaps on real projects in our firms, introducing what surely must be efficiencies. Think on this idea. Where does it take you? What does your nirvana workflow look like?

 

Join the discussion at the Design Technology Summit next month and bring a napkin sketch of your nirvana!

I dug up an old workflow diagram of the schematic design phase from 2009 to help get your cogs turning:

Register your interest for DTS here

Knowledge Management, AI, and Mentorship

Are you as crazy about Google Maps as I am?  I really love being able to check it before leaving on a drive – if there’s traffic on my way, it will tell me.  If I need to reroute, I can avoid complications before I run in to them.  I can evaluate the best route, be on my way, and not worry.  Google will even correct mid-trip if it sees congestion and finds a better route.

 

In spite of my enthusiasm for Maps, I’m a little disappointed when I use it.  I find that I don’t recall the directions as well as I do when I have to figure out the route on my own.  I focus on obeying Google’s discrete, turn-by-turn directions at the expense of understanding the overall route.  In some ways I feel less prepared to make that trip again should my phone be unavailable, because I rely more on Google, and less on my brain.

 

Bringing this back to the topics at hand, I’m curious to see if there are parallels between my use of Maps and a person’s use of AI-assisted Knowledge Management.  While I’m always in favor of retaining knowledge, I hope we can find a way to do so that augments rather than compromises the learning experience.  How can we make information available in a way that stimulates learning, and reinforces the mentorship model?

 

As one of our topics for this year’s DTS, I hope you’ll bring some thoughts on this topic to DTS this July.  Let’s discuss!