by Gary Mintchell | Apr 28, 2021 | Capital Projects, Leadership
Yesterday, I sat in on a Webinar from Honeywell about a plant optimization project with Woodside. Here are a few takeaways.
Supplier/Customer Collaboration–from the earliest phase of the project, the customer brought in experts from the supplier to assist planning, specifying, scheduling, and the like.
Planning–not a surprise to any of us who have done any project in manufacturing (or around the house) that success was correlated with good planning.
Access to remote experts–we now have good tools for bringing in experts from wherever they are to consult with the project. Video tools mean they can see and be seen. This saves time, money, headaches.
by Gary Mintchell | Apr 27, 2021 | Business, Leadership
Jason Fried, co-founder/CEO of Basecamp a projects software company, released a blog post (the hyperlink on his name) that dumped a number of new policies on employees. He and co-founder CTO David Heinemeier Hansen (@DHH) have decided that employees at Basecamp are too worried about things other than work.
Following the lead of Bitcoin, they have banned all political and social communication on company communication tools. Employees are free to do that on their own time on their own social media platforms. But not employee-to-employee.
I have to back up a second to some of my past experience. I served eight years on a public school board. I learned that school administrators hate any public discussion and questioning of their decisions. They hate any feedback from teachers. Since principals are “part of the club”, it becomes career-limiting for a principal to question anything. I mentioned one time to the superintendent that I was advising a bunch of students on how to protest (thanks to my civil rights/hippie days). He blanched.
Similarly, Fried wrote that decisions wouldn’t be discussed. Live with it and go back to work.
They also did away with “paternalistic” policies. They had over time instituted policies and payments for wellness programs and the like. They will give employees a payment this year in lieu of the benefit, then it’s cut out. Maybe in future years a profit sharing plan will make up the difference. The rationale is that they don’t think the company should tell people what’s good for them–even if it is.
And, no more advisory committees. The person in charge makes the decision. Period. If they want feedback and information from anyone, they will ask for it.
Basecamp has been an employee-friendly company. These new policies require trust. They blew the trust by the way they rolled it out. Bitcoin lost a number of employees with its new policies. We’ll see how many Basecamp loses. And what the culture will evolve to. And whether Fried and DHH will write any more books about the right way to work.
by Gary Mintchell | Apr 27, 2021 | Automation, Software
I have witnessed the evolution of Optical Character Recognition (OCR) over the past 35 years. This is an automated system of taking a picture in a digital vision system of some text, doing some magic processing, and outputting machine understandable text that can be used directly in your software application.
Neurala discovered this website’s reach and has been sending me a stream of updates. This is a company moving forward rapidly. Today’s announcement pushes the state-of-the-art.
Today, Neurala announced a partnership with IHI Logistics & Machinery. Neurala’s vision AI software will be deployed to increase the effectiveness of optical character recognition (OCR) reading of package information by automatically identifying expiration dates, to ultimately reduce waste and relieve workers from mundane, repetitive tasks.
IHI Logistics & Machinery is a leading global provider of material handling and factory automation solutions, with a focus on the management of food packaging information and logistics process improvement specifically. Traditionally, food and perishable items come into the warehouse with a production and an expiration date, with these important dates scanned by human workers with handheld OCR terminals upon arrival. It is a tedious job, and when an OCR terminal misreads an expiration date, it results in the need for inspection by humans. This also increases manufacturers’ costs and reduces profits.
Neurala’s vision AI will improve OCR by automatically identifying a product’s expiration date, including validating where on the packaging the expiration date is located. It will also be able to verify that text on a box is the expiration date, as opposed to other numerical data such as the SKU or production date, if a series of dates is present. This reduces the need for manual intervention when errors or misreads occur and ensures that only accurate data is passed back to the ERP system.
“Introducing AI and automation into our workflow will be a game changer for our business,” said Takayuki Sado, general manager at IHI Logistics & Machinery. “By partnering with Neurala, we are able to bolster our value to our customers, by dramatically increasing the speed and efficiency of material handling. This level of automation is also extremely valuable, as it helps us do more with less – which is especially critical in a time when there are restrictions limiting the number of workers present on the warehouse floor.”
“Neurala is on a mission to help manufacturers realize the benefits of vision AI by partnering with companies around the world who are leaders in their industry,” said Max Versace, co-founder and CEO of Neurala. “We are excited to partner with IHI Logistics & Machinery to provide them with the technology needed to further their position as an innovator and leader in material handling.”
by Gary Mintchell | Apr 27, 2021 | Personal Development
That was a popular ad phrase in the mid-80s. It referenced the quality of the digital recording on cassette tapes. (Pre-CDs and pre-streaming.) Given the devolution of social media in the past few years, these days we ask, “Is it real, or is it AI?”
I got this cute and thought provoking website from a PR agency. I told her it looked as if people were not all that observant. But, that would have been unfair to people. Some images were impossible to tell. Try it and let me know. Then think before you leap on stuff you see on social media.
As Artificial Intelligence is one of the latest trends that advances so rapidly, we decided to conduct the experiment to see if internet users are able to recognize photos, artworks, music, and texts created by AI.
I believe these results will be fascinating and shocking also for your readers! Here you can find the whole data-based article:
Can you tell the difference between human and AI?
The significant findings of our Human vs. AI Test were surprising:
- In some survey groups, as many as 87% of respondents mistook an AI-generated image for a real photo of a person.
- Two-thirds of people who think they would recognize a chatbot struggle with identifying AI-generated texts.
- About 24% of male and only 17% of female respondents made less than 3 mistakes when judging photos.
- Interestingly, more than 37% of respondents who declared as non-binary made less than 3 mistakes when judging photos (out of 7 examples).
- The most crucial factor in the test score was familiarity with AI technology.
by Gary Mintchell | Apr 26, 2021 | Automation, Robots
This conversation came to me from the Hannover Messe Digital Event. It’s rather interesting in that these people market robotic solutions. They think, however, that sometimes (often?) companies implement high-end and sophisticated automation and robotics that are often too complicated to be used. So, they are turned off. I learned first-hand as a provider of automation technology when the user/operator does not understand or even fears the technology or when the user interface is too complicated then the system will be turned off and manual operations will be used.
These comments are worth considering when planning new projects.
Manufacturers in Europe struggle with new automation and robotisation solutions as the technologies they deploy do not provide the desired return on investment and often end up standing idle. Torsten Christensen, partner and co-founder at ChangeForce, a Danish industrial consultancy firm, states that this could be changed by shifting focus towards standardised robots in-house employees would be capable of tuning.
“Europe, as well as the rest of the world, is heading to a new record of operational industrial robot stock in 2020, but these raw numbers tell only one half of the story. In reality, I think the industrial robotics industry is experiencing something we witnessed some 20-25 years ago in the dot-com era: lots of hype coupled with a low return on investment, heavy reliance on third-party integrators and, quite often, complex machines not returning the expected value. I believe European manufacturers, first and foremost the SMEs, should take one step back and rethink their robotisation strategies to avoid even larger disappointments,” T. Christensen says.
He is seconded by Thomas Ronlev, the CEO at Factobotics, a Danish-Lithuanian maker of standard industrial robots, presenting RoboBend, the world’s first standard sheet metal bending robot, and Flexy-Weld, a unique flexible robot welding solution, at Hannover Messe Digital Edition this week.
Th. Ronlev says that the risk of robotisation failures further exacerbated by a very high current robotic solutions cost. His company aims at traditional industrial processes and custom situations that can be scalable across a broader array of manufacturers.
“European manufacturers often believe their production processes are unique. They are also often convinced that entire production and supply chains be automated. I think that truth is much more down to earth: some processes are indeed ripe for automation, but not all of them, let alone at once. We would see a much higher degree of satisfaction if companies started from simple processes and standard robotic solutions that come at a fraction of the cost and are easy to master. Some of the most successful factories use relatively simple robotics,” Th. Ronlev says.
Factobotics has been demonstrating its solutions tackling these challenges at HM Digital Edition this week. RoboBend robot solves the problem of finding qualified machine operators, provides higher capacity on the company’s present machines, lowers production costs and consistently delivers high quality. RoboBend is designed to make it simple to use at any production environment, for any worker with no special training.
The company has also been demonstrating Flexy-Weld, a unique flexible robot welding solution that is currently in development and field-testing phase in Lithuanian metal processing company LT Technologies. This solution, based on an innovative flexible hexapod technology FlexHex that adapts to any new piece and holds it in place for welding, completely eliminates the need for jigs/fixtures. Flexy-Weld also dramatically increases flexibility and agility of production, eliminates the need for hundreds of different jigs/fixtures (warehousing, time waste), therefore increasing capacity and productivity, saving time and increasing earnings.. Attached are a few pictures, too.
