Sam: One Robot, a Dozen Engineers, and the Race to Revolutionize the Way We Build

by Jonathan Waldman

A true story of innovation, centered on a scrappy team of engineers--far from the Silicon Valley limelight--and their quest to achieve a surprisingly difficult technological feat: building a robot that can lay bricks.Humans have landed men on the moon, programmed cars to drive themselves, and put the knowledge of our entire civilization in your back pocket. But no one--from MIT nerds to Army Corps engineers--has ever built a robot that can lay bricks as well as a mason. Unlike the controlled conditions of a factory line, where robots are now ubiquitous, no two construction sites are alike, and a day's work involves countless variables--bricks that range in size and quality, temperamental mortar mixes, uneven terrain, fickle weather, and moody foremen. Twenty-five years ago, on a challenging construction job in Syracuse, architect Nate Podkaminer had a vision of a future full of efficient, automated machines that freed men from the repetitive, toilsome burden of laying bricks. (Bricklayers lift the equivalent of a Ford truck every few days.) Offhandedly, he mentioned the idea to his daughter's boyfriend, and after some inspired scheming, the architect and engineer--soon to be in-laws--cofounded a humble startup called Construction Robotics. Working out of a small trailer, they recruited a boldly unconventional team of engineers to build the Semi-Automated Mason: SAM. In classic American tradition, a small, unlikely, and eccentric family-run startup sought to reimagine the behemoth $10 billion construction industry--the second biggest industry in America--in bootstrap fashion. In the tradition of Tracy Kidder's The Soul of a New Machine, SAM unfolds as an engineering drama, full of trials and setbacks, heated showdowns between meticulous scientists and brash bricklayers (and their even more opinionated union), and hard-earned milestone achievements. Jonathan Waldman, acclaimed author of Rust, brings readers inside the world of the renegade company revolutionizing the most traditional trade.

Genres: Nonfiction, Business, Science, Engineering

1 pages, Audio CD

First published January 7, 2020

About the author

Jonathan Waldman studied writing at Dartmouth and Boston University's Knight Center for Science Journalism, and worked in print, radio, and TV before landing in books. His first book, Rust: The Longest War, was a finalist for the Los Angeles Times Book Prize and winner of the Colorado Book Award. His writing has otherwise appeared in The New York Times and McSweeney's. Visit him at JonnyWaldman.com.

Reviews

[Brad Feld · Rating 5 out of 5]

Jonny has delved deep into an intriguing entrepreneurial story that includes robots, bricks, construction, and all of the ups and downs of trying to get a business off the ground. Unlike many of today's 'just raise a ton of money and then good things will happen' stories, this is an awesome tale of grit, endurance, and relentless optimism against almost all odds. When you toss in the history and evolution of a bricklayer, you get a fascinating story. Read it—you'll love it like I did."

[Lori · Rating 4 out of 5]

Amazing that everyone in this company didn’t quit. Check out the video on CR’s website. A construction site reality show. I don’t know how they handle the frustration and politics but they did, and letting technology do the heavy lifting is very cool.
Congrats to them all. Good story. This is what is best about America. People with dreams who don’t give up, can’t do the 9 to 5 with incompetent bosses, and have to see their vision realized. We need more people and more stories like this.

[David Wineberg · Rating 3 out of 5]

SAM is not the story of the future of automation or the power of artificial intelligence to disrupt life as we know it. It is instead the biography of a startup totally focused on developing a machine to build walls of brick. SAM stands for semi-automated mason.

In Jonathan Waldman’s telling, it is a long slog of steps, generally two forward and one backward. The union did not want automation. Companies didn’t see the potential. Bricks were being replaced by blocks and panels. And the whole construction industry was distrustful when it wasn’t simply skeptical. Nothing new in any of that; it’s what every startup’s sales team faces daily. And hundreds of books have been written about it.

Waldman follows the activity in microscopic detail. Every employee gets deeply profiled, from his (there are no women) school interests, to career path and annoying habits at work. There are so many players it is difficult to remember who is who and what is so special about them when they reappear later.

The book alternates between progress at the company and diverting chapters on things like the history of mortar and the history of the masons union in the USA. There’s even a long biography of the CEO’s high school swimming coach, a famous professional wrestler called the Destroyer, including all the big name wrestlers he fought, his tour of Japan, and his ending up coaching young swimmers.

The SAM machine itself goes through an evolutionary process, just as in any other startup. Theory got tested in the field with real customers. Unlike many startups, Construction Robotics seems to have had an endless supply of firms willing to do trials on actual jobs of theirs. SAM kept failing then improving, and failing in new ways. There were unconsummated deals for sales, near-death crises, a constant lack of sufficient capital, an angel footing the deficits, and a secondary product that apparently saved the firm from liquidation. The biggest problem, or at least the one most recounted in the book, is faulty wi-fi between the control tablet and SAM, laying bricks up high on some wall.

The book reads like an official biography. It’s all good and ends well. There’s a positive spin on just about everything. It is an immortalization of this company, for future leaders to read and be inspired by. What it is not is gripping. It is pretty much a straight line narration to eventual success, albeit not with the original goal and product. There is no wider implication, no leverage and no suspense. The entire story is completely predictable, from engineers rushing out to sites to patch a part back into service, to big deals that could have made the company simply disappearing without any reason. It’s the same startup life we’ve seen a thousand different ways. I have lived it myself in at least three companies.

There is a lot of construction jargon which can leave the average reader flummoxed. Waldman also has chosen to employ non alphanumeric symbols for endnotes, which grow in length as the chapters add up. So it is not uncommon to see something like ¶¶¶¶¶¶¶¶¶ sprawling across a page in chapter 8, followed by, say §§§§§§§§§ instead of a single digit number.

SAM the machine has not revolutionized the construction industry, and SAM the book is not captivating.

David Wineberg

[Fen Kuntz · Rating 2 out of 5]

About 50 pages of interesting information stretched to 266. Too many chapters chronicling the boring daily jobs of people troubleshooting tech.

[Charles Haywood · Rating 3 out of 5]

The classic American path to technological success has been for driven tinkerers to obsessively work to solve a problem, from Eli Whitney to Thomas Edison to Steve Jobs. Such men strove to enrich themselves while benefiting others. SAM, the tale of one Scott Peters and his ten-year attempt to create a bricklaying robot, narrates such a story. True, his attempt was mostly unsuccessful, but then, most such attempts are. And in modern America, when excellence and achievement have been traded in wholesale for less-than-worthless “diversity and inclusion,” his is an inspiring tale.

Why did I read this obscure book? Masonry, like metalworking, has long fascinated me, and I have dabbled in both. Automation also interests me, for its economic and political effects. As I’ve written elsewhere, I think automation as a disruptive force is grossly exaggerated, across the political spectrum. This exaggeration has implications for, and says much about, our approach to progress today. The combination of these two topics, not usually tied together, grabbed my attention, so here we are.

A very old trade, not changed much in technique from hundreds or even thousands of years ago, bricklaying is a tremendously physically demanding task, for which, over decades, the mason pays a steep price. All the older masons I know suffer from a variety of physical debilities caused by their trade; years spent lifting heavy brick and block, moving and twisting, invariably take their toll. It is no wonder that the dream of a bricklaying machine has existed for more than a hundred years, and many men have falsely claimed to have created one. Peters decided he could be the first to actually do so. He wasn’t a bricklayer, though he had spent a lot of time on upstate New York building sites, which probably made him think the problem was easier to solve than it ultimately proved to be. But you don’t know until you try.

Why has automated bricklaying always been an intractable problem? For the same reason automation has, so far, proved impossible for the vast majority of human tasks: complexity, much of it not obvious to the outsider. There are a nearly infinite number of variables involved in bricklaying, that combine in myriad ways, all of which an experienced human can easily process and synthesize, but which have to be specifically identified and precisely programmed into a machine for it to do anything at all. Although it is not the author’s goal, this book shows the impossibility of much automation, contrary to the glossy promises sold by and to the professional-managerial elite who have never worked with their hands or tried to understand anything physical, other than paper-pushing at a BS job.

I have personal experience with attempting to roboticize an industrial process. In fact, ten years ago, I intended to almost entirely roboticize my own production facility. Essentially, my factory mixes and fills a variety of foods into containers, a complex set of steps different for every product. After discussion with many of the largest robot manufacturers, such as Fanuc, I was assured that robots could be designed that would assemble and fill containers, then seal and label them, all at high speed without any but occasional human oversight. Moreover, I was assured this could be done for a reasonable, even cheap, price, with a short payback period. This was attractive, because it would save on labor costs, and since my operation is short-run and flexible, filling many different products into many different types of containers, it is a high-touch business with high labor costs.

Every few years extensive press is given to some new development that will supposedly make robots ubiquitous, but never does—such as Baxter, the pick-and-place robot with the emotive face made by Rethink Robotics, which soon went out of business. That should have told me something. None of my competitors used robots, and I should have known that there was a reason for that. Like all dreamers, though, I thought I had seen something others had missed. My conclusion from that industrial robots were very rare outside of the automotive and distribution industries was that I could be the pioneer! So I took the plunge, and bought a robot. I was promised it could take a constant stream of jumbled jars and lids, identify and pick them, fill packages quickly and accurately, seal them, and send them on down the line, allowing me to mint the money I deserved.

It couldn’t do any of that. It’s possible that a robot could be designed, for many millions of dollars, to actually do that—but at a fraction of the speed a traditional set of filling machines could do it, at twenty or thirty times the cost. In practice, my robot could do nothing accurately or at more than a snail’s pace, and required constant attention and service from a legion of human beings. Why? Because robots, and automation in general, can’t handle variability. They excel, as this book shows, only at repetitive tasks with variables that can be narrowly specified, and with no unexpected variations or random occurrences. But in my industry there are already finely honed machines for the tasks I needed done, machines that can be quickly and easily adjusted by human operators to deal with variability, based on their experience. So in my industry, and in most industries, robots are worthless, adding huge expense but finding no outlet for what their strengths are—the ability to be programmed with variables that can be precisely quantified, and to move in multiple dimensions faster and more accurately than a human can, or in toxic environments dangerous for humans.

And this book, despite much inspirational talk, shows this same variability, and thus the uselessness of robots, is true in bricklaying. Bricklaying variables include a kaleidoscopic variety of cross-interacting environmental factors (rain, wind, sun), variability in materials (brick size, mortar rheology), and variations in measurements as built from plans. All of these a mason, after long experience, can address on the fly, but they choke a machine. The same is true for nearly any industrial process where conditions cannot be precisely specified, meaning most industrial processes. Thus, welding robots make sense, because where the variables are few, the material’s condition and location in space can be precisely specified, and exact quality parameters are easy to determine. Bricklaying robots, where none of that is true, don’t make sense.

But how can this be? We are always told that automation of everything is just around the corner. Proponents of automation, recently most notably Andrew Yang, casting themselves as mere realists, wave their hands at the truth that robots can do little, promising that artificial general intelligence is imminent, which will allow robots to synthesize variables just like a human being. However, there is zero evidence of such AI being possible, which is why autonomous cars will never be used outside of controlled test environments, and fifty years from now men will still be laying bricks by hand. But we shrink from admitting that we can’t accomplish great technological things like our forefathers, so we preen ourselves with the fantasy that we can through artificial intelligence, while in fact accomplishing nothing except mass manufacture, using slave Chinese labor, of shiny baubles, sold to consumers to rot their minds and allow them faster access to more personalized pornography. Whatever progress is, that’s not it.

So, no surprise to a reader who understands these truths, SAM (“semi-automated mason”) was not a success. The book, though published in 2020, ends in 2016, suggesting that SAM was gaining acceptance in the marketplace. You have to do a little research on the company profiled, Construction Robotics, to realize this is untrue, because the author, Jonathan Waldman, never quite comes out and says it. But after 250 pages chronicling in great detail the ups-and-downs, mostly downs, of the small group of men trying to build SAM, you realize that they failed. There is little recent mention of the machine on the internet, and if you go to the website for Construction Robotics, while SAM is mentioned, their main product appears to be MULE (“Material Unit Lift Enhancer”), a device that is not a robot at all, but a flexible and clever lift-assist device for helping a mason lay concrete block without breaking his back. No new iteration of SAM has been brought out for five years. It’s quite apparent that SAM is not going to change the bricklaying industry. Yet that is the fate of most men who obsessively work to bring a dream to life; it is not a criticism of Peters, who at least makes machines that help society and add real value, unlike the vast majority of output from Silicon Valley.

The story is good, but the book is not. It is not a pleasure to read, even for someone interested in bricklaying or the building trades. It’s mostly just a sequential description of hard-to-distinguish vignettes of the Construction Robotics team trying to make SAM work on various masonry jobsites. It offers zero pictures, which makes the book much less interesting, and also makes it impossible to keep track of who is who, especially with the high turnover at Peters’s company. The lack of an index also frustrates the reader, and there are red herrings that seem important, but aren’t—such as constant references to a quarter-million-dollar Leica laser system that would supposedly solve all SAM’s problems, which I expected to show up for testing some time, but instead just stopped being mentioned near the end of the book. This book might have made an interesting Medium article, but that’s about it.

Why are we constantly promised innovations that never arrive? There is little doubt that we are in an age of technological stagnation in every area, despite the flash and ubiquity of mobile phones and similar geegaws. I am not concerned here with why they don’t arrive—I know that already, corruption and decadence. (If you want to cry, go read American Genesis, by Thomas Hughes, a 1989 book that chronicled a brilliant past yet did not realize the future was stupid. To be sure, Hughes notes what others, including Mariana Mazzucato, have also pointed out—that for at least a hundred years, most great advances have been team efforts, though usually driven by the vision of one man, and often receive substantial government support, as they should.) What I want to know is why false promises denying the obvious are constantly made to us. This is true in every area of life—read any news site, and we are told complete lies about everything from autonomous cars and trucks to cures for cancer. And that’s ignoring the “news” of technological developments that is simply scams, like “green energy” and “green industry.” Why?

It must be a combination of the ruling classes being unwilling to admit their rule has created a failed society, despite their utopian promises of a remade ideal society, and the (tightly connected) dominance of stupidity and gullibility among the masses, as shown in popular culture—the “I’ll buy that for a quarter!” of Cyril Kornbluth’s classic story “The Marching Morons” (parodied in the movie Robocop). Both groups like to believe the fantasy that the technological future is bright and we will achieve more than our fathers did. Oh, some of the former group know it’s a lie, but it’s an instrumental lie, a type of consumerism that keeps the masses quiescent, or at least they hope it does. Most of the former group, who indoctrinate the latter group, don’t know it’s a lie, and can’t wrap their mind around the possibility, because they do not want to believe what used to be a commonplace—that success and progress require sacrifice and differentiation that depend upon and reinforce hierarchy. They instead believe in their bones everyone is both equal and above average, and that emancipation means everyone can be Wernher von Braun. They believe the fountain of progress dispenses its benefits to anyone who cares to drink, and that the incompetent and untalented are just as able to achieve as the obsessed geniuses who in the past drove achievement (and that past geniuses were overrated, guided by hidden figures who were the real talent).

Examples of such lies are everywhere. Let’s take something that’s not robots. Let’s take nuclear reactors. As we all know, the nuclear power industry, originally seen as able to create power too cheap to bother metering, foundered decades ago on the rocks of hysterical fear spun out of environmentalist fever dreams, combined with choking government regulation. Quite often, though, we are told that just around the corner is a new Nuclear Age that will solve all our power problems. Yesterday, for example, prominent people on Twitter (e.g., Jack Dorsey and Scott Adams) noted with excitement that “America Just Made a Huge Investment in Next-Gen Nuclear Power.” If you read the details, though, it’s silly. A modest government budget has been set to build smaller reactors based on current technology. And are they imminent? No. Any actually new reactors are “strictly in the realm of the imagination in 2020, but will be ready for the runway by the ‘mid-2030s.’ ” In other words, they will never arrive, and the money will be dissipated among grifters. Other reactor technologies, such as thorium reactors, have been pushed for decades, but they will never arrive either. Our sclerosis is too far advanced. Yet we hear, constantly, how bright the future of nuclear power is, a fantasy one step less insane than believing we will have weekend trips to Saturn this century.

So our society is technologically defunct, or, more accurately, decadent. What about . . . [Review completes as first comment.]

[Amanda · Rating 3 out of 5]

This was pretty dry and left me dissatisfied. I was left wishing more topics were explained while others topics were overly detailed and unneeded. The transitions felt disjointed when other topics were mentioned. Knowing the family history of one person and then jumping back in the story felt like commercial break whip lash.

[Quentin Baker · Rating 4 out of 5]

A clear, nuts-and-bolts look into the problems of robotics and, maybe more importantly, the difficulty of introducing automation to one of the oldest professions, with all its layers of tradition and expectations.

[Vishnu R · Rating 5 out of 5]

It was fun for me to read this book. Very authentic narrative of the struggles of a scrappy startup.

[Bouke · Rating 2 out of 5]

Not enough technical details on the robot and how it works, too much on all the people involved.

[Charlene · Rating 5 out of 5]

I wish there had been a little more discussion of the brainstorming that went into fixing the problems and bugs within the robot.

[David V. · Rating 4 out of 5]

Received as an ARC from the publisher. Started 11-6-19. Finished 11-9-19. One of the more unusual books I've read. Think the idea of a robotic brick-laying machine is dull, think again. The research and development of just such an actual machine was fascinating. This would make a wonderful movie in the vein of "Tucker: The Man and His Dream." It should be read by anyone interested in robotics (even students in schools); construction workers; and R&D specialists. It also shows the old saying--if it can go wrong, it will---but perseverance won out for these developers. It took years but they did produce a functioning brick-laying robot.

[Kim]

A book for sesquipedalians. The fifty cent words made the prose too clunky to engage with for me.

[Leah T · Rating 1 out of 5]

Read for a class as a case study on how not to design robots lol