Robotics and Automation
Article | July 13, 2022
Feature engineering is the process of extracting new variables by transforming raw data to improve the predictability of a machine learning model. But feature engineering is not just this kind of simple translation of categories like names or colors into numbers. The following section includes a collection of different kinds of engineering approaches that go beyond a translation of categories into numbers and address needs such as transforming numbers into categories or filtering data points due to missing or false data.
Article | July 20, 2022
640K Ought To Be Enough for Anybody
That infamous statement—whether Bill Gates said it or not, goes to show the change in computing and the demands on the semiconductor industry over the last 40 years. At the beginning of the ‘70s, there was no expectation that the personal computer could become an affordable item for the man in the street. By 1979, however, Atari had released the 400 and 800 series of home computers. Three years later, the Commodore 64 made its debut, featuring 64KB of RAM and using an 8-bit CPU.
In 1977 Steve Wozniak designed the Apple II, an 8-bit home computer. Launched at the 1977 West Coast Computer Faire, it was aimed at the home consumer market rather than the business market.
August 12 1981: The IBM PC
It could be argued that the first non-Apple PC, as we've come to know it, was the IBM 5150 personal computer. Its success spurred the production of IBM clones, or IBM PC compatible computers, with Columbia Data Products (CDP) producing the first in June 1982.
A Case of History Repeating Itself?
In the early ‘80s, anticipating the demand for PC’s to continue, memory chip manufacturers ramped up the production of RAM. But by September 1985, the market had stagnated, and a DRAM chip could be bought for $2.95. Demand for computers had slumped, and this low price reflected industry slowdown and extreme overproduction.
Roll On To 1988 and the Price of Computer Chips Rocketed
With a glut of existing RAM chips in the marketplace, manufacturers were cautious of the overproduction of 256-kilobit DRAM chips and converted their factories for 1-megabit chips.
This shows that chip producers drastically misread the market. 1-megabit chips took significantly longer to manufacture, and so before too long, there was a shortage of RAM, causing prices to increase.
The situation began to improve by the next year. Since then, although there have been years when supply was affected, it has been nowhere as catastrophic as now.
The Current Semiconductor Chip Crisis
We are facing an unprecedented shortage of semiconductor chips which is affecting worldwide markets. COVID-19 undoubtedly had a massive influence on this, but the demand for microchips was already soaring.
2019 COVID-19 Appears
Although in November of 2019, a person displayed the first detectable case of COVID-19 in China, there was little else to warn of what was about to come. In the following months, as cases increased, so did hospital admissions. With the horror of widespread deaths, we saw countries bringing in protective measures and restrictions. These became increasingly severe and ranged from social distancing to working from home. This had a direct effect on the industry, seeing output slowed or temporarily ceased.
In February of 2020, the indirect effects of the pandemic began to bite. Companies closed offices or limited the number of onsite staff. Employees were also encouraged to work offsite where possible. Other employers were forced to furlough employees. For some, the concept of working from the home kitchen table became a daily reality. This was seen as essential to ensure that services were maintained, albeit at a reduced capacity.
For factory-based and hospitality industries, the impact was more dramatic.
Company Employees Working From Home
The decision by many companies to encourage staff to work from home was a direct result of COVID-19 and the resulting Government restrictions, but this, in turn, caused an indirect effect on the semiconductor shortage.
In some cases, employees might have had existing company laptops issued to them for use in the workplace. In other cases, the use of their personally-owned device might have been sanctioned for company work. But in other circumstances, the company would either pay the employee to upgrade their laptop or provide a new company laptop with the necessary application and security software installed.
This added to the general increased demand for computers that had computer manufacturers struggling to procure chips.
Many furloughed employees suddenly finding themselves trapped at home with limited opportunities (if any) for socializing turned to or spent more time on gaming.
This fuelled an interest in the latest products on the market and a keen appetite for products about to be launched. In turn, manufacturers clamored for more chips.
Schools and colleges rose to the challenge of providing continuing education for their pupils and turned to online teaching when the school buildings were closed. Children were being home-schooled by parents and following online lessons. But it was essential that the children had the necessary resources. The basic requirements were a laptop with a webcam and a reliable internet connection. Subsequently, laptop sales increased dramatically.
Who Is Taking The Hit?
A simple answer is— any industry whose products depend on a high level of semiconductor chips, but in particular, the main players feeling the pain are the automotive and consumer electronics markets.
The Auto Industry
Automotive Companies Fall To the Back of the Semiconductor Chip Queue
As COVID-19 began to take a grip, and with falling demand for vehicles, auto manufacturers either closed sites temporarily or reduced operations. Subsequently, they scaled down backorders from semiconductor suppliers. Meanwhile, the consumer electronics market was thriving and crying out to suppliers for more semiconductor chips.
Later, when manufacturing was resumed, auto manufactures found themselves at the back of the queue.
What Chips Are Used In Motor Vehicles?
There are various types of chips used by auto manufacturers in their vehicles, ranging from commodity chips to microprocessors.
According to Statista, “Infineon, NXP, and Renesas were the leading automotive semiconductor manufacturers worldwide in 2020. Infineon's market share was estimated at around 13.2 percent. The total market in 2020 was sized at around 35 billion U.S. dollars.”
The Domestic Market (Consumer Electronics)
Broadly speaking, this sector covers anything that falls into the entertainment, communications, and recreation categories.
Although visits to high street stores to make purchases proved difficult, if not impossible during lockdown periods, online sales soared. But this boom has caused manufacturers a headache, as launches of new products have had to be delayed and fulfilment of the demand for existing models could not be met due to the chip shortage.
Other Contributory Factors to the Crisis
Although COVID-19 disrupted chip manufacture by causing foundry shutdowns and the halting of production, it wasn’t the only factor. An already beleaguered market was battered by other factors compounding the chip shortage crisis.
Drought in Taiwan
Water, a major necessity for semiconductors production, has been in short supply due to the worst drought in 56 years.
Suez Canal Blockage
In March 2021, the 400-metre-long (1,300ft) container ship ‘Ever Given’ ran aground in the Suez Canal and blocked the channel for six days, further impacting distribution and supply.
Japanese chipmaker Renesas Electronics Corp. the world's third-largest supplier of automotive chips suffered a fire at its factory.
Severe Weather Conditions in Texas In February
Samsung, NXP, and Infineon chip fabs shut down in Texas amid record storm.
Why Not Just Produce More Chips?
While attempting to address the global chip shortage as expeditiously as possible, semiconductor manufacturers cannot afford to make a knee-jerk reaction. If fabrication plants are at maximum capacity or are only structured to make one type of chip, why not build more fabs?
Semiconductor wafer fabs are hugely expensive to build. It takes considerable time to construct a new fab, with some as large as small cities. These fabrication plants, also known as foundries, require highly controlled environments where temperature, humidity and static electricity are controlled, and dust-free environments are guaranteed. As an immediate response, building new fabs is not a practical solution to the problem. Long term strategies will have to be put in place as the whole situation is addressed.
When Will The Global Chip Shortage End?
There are differing views being expressed on this tricky question. Some are optimistic, considering that the worst of the situation is over. Others provide a gloomier outlook, warning that we could be experiencing shortages well into another two years.
Crisis management expert Edward Segal writing in Forbes: “The semiconductor chip crisis that hit companies around the world shows no signs of ending any time soon and will continue to impact the supply chains for many industries. Indeed, some organizations have yet to fully recover from the impact of the blockage of the Suez Canal last March on their ability to send and receive essential materials, parts and supplies.”
Are There Any Lessons To Be Learned?
A cynical reply might be—expect the unexpected.
Of course, it is impossible to predict and plan for every possible eventuality. Changing market trends should be anticipated, whereas something as unforeseen as a global pandemic cannot. Manufacturers, however, should seriously take a look at their contingency plans.
It seems that far and above the other problems of the chip crisis, the biggest headache within the semiconductor industry is the supply chains.
Writing in an article for ZDNet, Daphne Leprince-Ringuet: “The semiconductor supply chain is flawed, and it's going to take a long time until things get better, despite the combined efforts of industry and regulators.”
Supply chains are the highways of trade upon which product delivery depends. But the semiconductor supply chain is hugely complex and is spread across several countries. Admittedly, it is essential to create more fabs over the following years. Still, it is critical to maintain a watchful eye on supply chain policies to ensure future semiconductor chip demand fulfilment.
Just-In-Time (JIT) Model
Considered as an effective approach by some automotive manufacturers as an efficient method of business management in times of plenty. It becomes counterproductive in times of shortage when they will face long chip manufacture lead times.
Chip manufacturers are advocating a greater knowledge of their customers’ production maps, stating that even a two quarter indication is insufficient for planning.
Think Outside The Box
Be open to some lateral thinking. Recycling could be an interim response to chip shortage.
All Of A Sudden Vintage Equipment Is A Hot Commodity
Steven Zhou writing in Forbes, reports that old (obsolete) fabs could be suitable for the production of some current 'smart' devices.
While the creation of extra fabs can take over two years and the building of the manufacturing equipment up to eighteen months, repurposing old equipment could be a source of additional capacity.
Reliance On Asia
The current crisis has brought about an awareness of the inadvisability of an ongoing reliance on Asian fabs for the supply of semiconductor chips for U.S. and European markets.
According to the Semiconductor Industry Association (SIA) in a publication Strengthening The Global Semiconductor Supply Chain In An Uncertain Era “Over the next ten years, the industry will need to invest about $3 trillion in R&D and capital expenditure globally across the value chain in order to meet the increasing demand for semiconductors.”
Moore’s Law Is Not Dead
Moore's law is the premise first expressed in 1965 by Gordon E. Moore, the co-founder of Intel, that the number of transistors on a microchip doubles every two years, though the cost of computers is halved. Or put another way—that we can expect to see larger-scale integration with more circuitry packed into chips for the same form factor.
If this proves true, manufacturers will take advantage of these cheaper and more advanced chips to develop a new generation of products that consumers will be only too eager to buy.
Article | August 25, 2021
Since we first learned to record sounds on wax tubes that could be played back on phonographs, humanity has come a long way. While setting up one of these antique music players and hearing what audio enthusiasts listened to in the late 1800s is interesting, it pales in comparison to the advances we've made. What are some of the most exciting recent acoustic engineering advances, and what does the future of acoustic engineering look like?
Recent Audio Engineering Advances
Inflatable concert halls have appeared in the most unexpected places, providing sound engineers with the perfect canvas to create the ideal stage for musical acts. Once the structure has been inflated and is ready to go, sound engineers can manipulate the interior acoustics in the same way.
Acoustic engineers can use 3D impulse response management to create a 3D visualization of a room's acoustics using a set of microphones. This makes it easier to identify problem areas that would otherwise be difficult to identify with the human ear.
Adding augmented or virtual reality to the mix introduces the practice of acoustic holography, which allows engineers to create a picture of the sound patterns in real-time. This could be useful for engineers concerned about noise pollution as well as those attempting to create the ideal acoustics for a musical performance or play.
What Does the Future Hold for Acoustical Engineering?
It's exciting to see how far the acoustical engineering industry has come, and even more exciting to see what the future may hold. What do acoustical engineers hope to achieve?
Although many people are unfamiliar with the term "acoustic engineering," it is a type of engineering that most people are likely to encounter on a daily basis. The future of acoustical engineering is bright, and we'll most likely see a lot of changes that make it easier to immerse ourselves in the perfect audio experience over the next few decades.
Article | July 21, 2021
A typical rebar detailing office has a spread of different ages working there – from those with just a few years of experience, to staff who have been detailing rebar for decades. In many ways, this mirrors the architecture, engineering, and construction (AEC) industry itself. However, as experienced staff are so quick to produce rebar documentation, often existing ways of working remain in use long after they have been replaced with more efficient methods. After all, by using familiar workflows only new detailers require training, rather than a whole office. Yet 3D rebar detailing software provides more benefits than just quicker deliverables – the benefits to the business, as well as the entire workflow, are significant. With the right rebar CAD software, this does not have to involve a steep learning curve that staff are not ready for, either.