So That Happened: Another Tesla Recall, AI for Manufacturing and USMCA Violations?
Editor’s note: Welcome to So That Happened, our editors’ takes on things going on in the manufacturing world that deserve some extra attention. This will appear regularly in the Member’s Only section of the site.
Tesla Model Y Vehicles Recalled… Again
Tesla has recalled 3,470 Model Y vehicles due to issues with bolts in the seat back frames that may have not been torqued to specifications in the second row, according to the recall report submitted at the end of February; the recall specifically affects 2022-2023 Model Y vehicles. According to the report, the estimated percentage with the defect is 4%.
“As of February 23, 2023, Tesla has identified 5 warranty claims, received between December 9, 2022, and February 14, 2023, that may be related to the conditions described above. Tesla is not aware of any injuries or deaths that may be related to such conditions,” according to the chronology section of the report.
This recall comes mere weeks after the company recalled over 360,000 vehicles due to driver-assistance technology issues.
Manufacturing Holes in the AI Hype
I hope you're enjoying all the hype about ChatGPT. I am (prepare to be shocked) cynical about the technology.
Artificial intelligence and machine learning (AI/ML) are good for data analysis. Trust an advanced amalgam of 1s and 0s to handle well a bunch of data that are also amalgams of 1s and 0s.
The technology will in the long run probably find good homes in Tier 1 customer service and IT help desk applications. “Did you turn it off and on again?” is easy for an AI to spit out a million times a day (and Roy will be grateful).
At my previous Endeavor Business Media outlet, Vision Systems Design, I wrote often about artificial intelligence and machine learning employed for inspection applications in manufacturing, making sure screws are in the correct position or parts meet specs, what today amounts to simple image analysis.
Manufacturers have, in this way, for decades shown what AI/ML is actually good for. These applications just don't make for sexy headlines in the mainstream press.
Not like Microsoft ‘lobotomizing’ its AI chat agent Sydney for demonstrating scary behavior; or finding out AI can make up answers out of whole cloth when asked a question; or discovering that these oh-so-advanced AIs might actually be stupid.
Anyway, back to manufacturing.
Rackspace Technology, a cloud server provider, in late February issued a report about how manufacturers are using AI and there were a few nuggets I found interesting.
Only 37% of respondents said they strongly trusted AI/ML with results. 35% strongly thought there are enough checks and balances to avoid the negative consequences of AI/ML. And only 35% strongly think there is sufficient governance in place to safeguard against any use of AI/ML. Click here for a lengthy infographic showing Rackspace's results.
I am omitting the “slightly think” results a) Out of my own bias and b) Because when it comes to AI, I think strong trust, secure trust, is all that really matters.
This is technology people want to use to drive their children to school (we assume autonomous cars will eventually be put to this use by parents that want an extra half hour in their weekday schedule), or help fight our wars, or care for the elderly.
There’s no middle ground on this. Either we trust AI, or we don’t.
So far, we shouldn’t. And according to the report, even if 42% of the manufacturers polled have realized substantial benefits and 35% have seen modest benefits from using AI/ML, 65% of the manufacturing IT leaders polled reported pushback or scrutiny about the penetration of AI/ML in their organization.
That number probably ought to be higher, in my opinion. Skynet jokes aside, this is technology that needs close supervision if we allow it to make decisions that meaningfully effect carbon-based life.
New Allegation Arises of Workers’ Rights Violation Under USMCA
The U.S. has called forth the USMCA’s labor dispute mechanism for the second time in 2023 after receiving a petition that workers’ rights of free association and collective bargaining were being denied at an automotive components plant in Mexico.
It was just last month that we shared with you a decision by the U.S. to invoke the United States-Mexico-Canada Agreement's Rapid Response Labor Mechanism to address allegations that rights of free association and collective bargaining were being denied at a Manufacturas VU plant in Piedras Negas.
This new assertion relates to a Unique Fabricating facility in Santiago de Queretaro in the state of Queretaro (part of the Bahia region in central Mexico, an area that some call the Detroit of Mexico for its large number of automotive plants and supplier facilities).
United States Trade Representative Katherine Tai has requested the Mexican government review the Unique Fabricating allegation, the Office of the USTR announced March 6.
“The USMCA’s Rapid Response Labor Mechanism is an important tool to defend workers’ rights and helps us advance a worker-centered trade policy around the world,” said Ambassador Katherine Tai.
This is the seventh time overall that the U.S. has invoked the USMCA’s Rapid Response Labor Mechanism, according to the USTR.
Instrumental Progress
Things are moving quickly in and around National Instruments Corp. these days. Less than two months after Emerson Electric Co. went public with an offer to pay more than $7 billion for the Austin-based company, executives have launched a full strategic review, reportedly narrowed down their list of serious bidders to three (including Emerson) and wrapped up their negotiations to add a German company to their portfolio.
Emerson executives’ announcement Jan. 17 that they are looking to snap up NI came after about eight months of on-and-off talks between the companies. In the seven weeks since, Reuters reported late last week, the NI team has talked to numerous other suitors and singled out Fortive Corp. and Keysight Technologies Inc. to join Emerson for a second round of bids. Reuters also said NI executives are targeting an early-April completion of takeover negotiations to cap their sprint toward a sale.
On top of that, NI President and CEO Lal Karsanbhai and his team on March 6 said they have paid an undisclosed amount for SET GmbH, a 22-year-old German company specializing in aerospace and defense test systems as well as semiconductor reliability systems in automotive applications. The two companies first linked up in late 2020, when NI took a minority stake in SET alongside Tech180 Inc.
The tools being brought to the table by SET, which has about 100 employees, will help scale NI’s work in the fast-growing electric vehicle space. As such, there’s a decent chance they’ll also factor into the second round of bids. Game on.
76 Million Engines and Counting
Back in January, we noted that Honda hit a major milestone, producing its 30 millionth vehicle in the U.S. in 40 years. General Motors’ Tonawanda Engine Plant in upstate New York looks at those numbers and responds, “Oh, how cute.”
The 1,300 workers at the plant make GM’s 4.3L V-6, 5.3L V-8 and 6.2L V-8 Ecotec3 engines for SUVs and pickups such as the Cadillac Escalade in Chevy Silverado. Oh, and the engine for the Chevy Corvette.
In 2020, GM announced a $70 million investment to expand machining operations in Tonawanda.
