What is Machine Learning (ML) and how does it work? How does it help in improving operational excellence in manufacturing and service industries? Machine learning is a sub-field of artificial intelligence and deals with a machine's capability to self-learn based on its surroundings and imitate intelligent human behavior. The basic concept is to create algorithms that learn from data and then apply them to new situations.

There are several types of machine learning, but broadly there are two types of machine learning: Supervised ML and Unsupervised ML.

Unsupervised machine learning aims to find patterns in unlabeled data, while supervised machine learning looks for trends in labels. In this post, I shall touch upon what is machine learning, where is machine learning used, benefits & challenges of machine learning and few examples of its industry applications. So, let’s get started…
 
Industrial uses of machine learning:
Machine learning is an innovation and is used in Industry 4.0 or smart factories. 

Machine learning is a vey powerful tool for business. It helps business leaders to predict customer demands, price products and segment customers. It also helps to forecast inventory levels and detect fraudulent activities. Machine learning’s capabilities range from image recognition to natural language processing. Businesses that rely on these technologies can provide better customer service, improve security, and even allow consumers to complete transactions without a human being. These innovations are changing the way we buy and sell goods.

Machine learning can be used to predict which products and services are best for a customer. Streaming service companies and social media apps use recommendation engines that analyze data to make recommendations based on personal preferences. Enterprises also use AI to predict customer churn. This allows them to better serve their customers and maximize profits.

Application of machine learning concept to technical and business processes improves operational excellence and in turn, increases profitability. 
 
Which industries will benefit most from machine learning?
Transportation industry for now. The transportation & logistics industry perhaps uses machine learning the most. Reinforcement machine learning trains models through trial and error and is often used to teach autonomous vehicles how to drive. It rewards the machine when it makes the right decisions.
 
Advantages and disadvantages of machine learning:
Some of the advantages of machine learning include the ability to customize education, which allows students to work on materials that are most suitable for their learning style. This can save teachers time because they do not have to plan a lesson for every student. Another benefit is an automatic grading system, which saves teachers time and gives a realistic assessment of student performance.
 
However, this technique requires a large amount of data to function properly. This is not only a time-consuming task, but it also requires resources such as CPU and GPU power. Furthermore, it may use more storage space than expected. In addition, machine learning algorithms can only be used as much as they have historical data to work with.
 
Another advantage of machine learning is that it allows for fast and automatic adaptation, which eliminates the need for human intervention. This feature is also used in anti-virus software and security software, which can automatically adapt to any situation without human intervention. The downside of machine learning is that it can sometimes make mistakes, causing inaccurate predictions or inaccurate advertising.
 
Challenges of machine learning:
Machine learning has a wide range of implications. Its use in making decisions can have a huge impact on the economy and our everyday lives. Research in this field is aimed at understanding the impacts and potential harms of machine-learning-based systems and developing tools and policies to avoid them. It also examines the optimal use of machine-learning systems to maximize aggregate social welfare.
 
Machine-learning specialists face many challenges in the deployment of their projects. Sometimes the algorithms they use are not accurate enough to solve a business problem. In this case, the project may fail. To avoid such scenarios, it is vital to have a team of experts with business qualifications working on the project. This will enable organizations to make the best use of this new technology and get faster ROL
 
In addition, organizations are under pressure to only develop machine-learning models that produce a positive ROI. The ROI of a model can depend on how users react, as well as how well it predicts the future. Moreover, it can be impacted by competition, public opinion, brand image, and algorithmic fairness.
 
Examples of machine learning in action:
Predictive analytics is one of the most exciting examples of machine learning in action. It can be used to make predictions about a wide variety of things, from real estate prices to product development. Machine learning algorithms use big data to make these predictions. They do this by analyzing billions of data samples and automating the process of annotating them.
 
Another great example of machine learning in action is the social media news feed. The recommendation engine in the news feed uses machine learning to recommend articles based on the user's past browsing behavior. This kind of learning is possible because the algorithms are able to improve their accuracy over time as they are used more.

Machine learning has found wide-spread applications in drug discovery & biomedical research, and improving operational excellence in financial institutions, healthcare, IT, transportation & logistics, retail & e-commerce, agriculture, aviation and education industry verticals. I have covered these in-depth in my book Machine Learning- How does it improve operational excellence in manufacturing and service industries. You may check out more about my book on Machine Learning here.


Related Reading:​

1. Continuous improvement tools
2. How to cut costs strategically using Kaizen
3. Streamline processes and workflows with Gemba Walk.
4. Top Ten Strategic Decision-Making Tools for Operational Excellence

​Follow Shruti on Twitter, Facebook, YouTube, LinkedIn

Categories:  Innovation | Operational Excellence |  Machine Learning

Keywords and Tags:
​#operationalexcellence #AI #ML #artificalintelligence #innovation #machinelearning #whatismachinelearning #machinelearningintrasportationindustry #profitability

Here's an exciting update on innovation in drug research!

The successful transplantation of human brain cells into a rodent's brain has made it possible for scientists to study the complex brain circuits and psychiatric disorders. The cells, known as organoids, are smaller than neurons grown in vitro. Researchers study these organoids to see how they integrate with the rodent's natural circuits.
 
Scientists have successfully implanted and integrated human brain cells into young rats' brains

In an exciting breakthrough, scientists have successfully implanted and integrated human brain tissue into the brains of young rats. The transplanted human cells did not cause any problems and were found to have no negative side effects. While this is an impressive accomplishment, it may not be the end of the story for humans.
 
The transplanted human neurons successfully teamed up with the rat's vascular system, which provides oxygen and nutrients to neurons. The scientists then performed an experiment to test how the human neurons reacted to external stimuli. They found that the implanted human neuron produced electrical activity when air passed across the rats' whiskers. This result suggests that the human neurons could influence the rat's behavior.
 
Organoids are smaller than neurons grown in vitro
 
Organoids, or cell cultures, are a relatively new technology. These small cultures of neurons have been used in the past to model aspects of human neocortical biology, including a variety of diseases. Organoids are particularly useful in studying infectious and genetic diseases of the brain. They can also be used to discover potential therapeutic reagents.
 
Organoids are similar to neurons in size and complexity, although they contain fewer neurons. Their cell lineage network is based on the same gene network as neurons grown in vivo. The cell lineage network for organoids and neurons is largely constant, except for variations in cell states during neurogenesis.
 
They integrate with rodents' own circuits
 
The rat-human hybrid brains are a ground-breaking advancement, offering new opportunities to study neurological disorders and raising ethical questions. These miniature human-brain-like structures can respond to environmental cues, allowing researchers to test therapies for disorders of the human brain. These brains were developed with the aim of studying the development and function of brain circuitry, including those that control behaviors and processes sensations.
 
The research involved genetically engineering human neurons and integrating them into the brain circuits of rats. The cells grew six-fold longer than those in petri dishes and were just as active as those in humans. They were able to send signals to nearby neurons and to distant ones, triggering the same responses in the rat brain.
 
They study complex psychiatric disorders
 
Scientists have successfully implanted and integrated a cluster of human brain cells into the brain of a newborn rat, an effort that could help them develop treatments for neuropsychiatric disorders. The goal of such experiments is to better understand complex psychiatric disorders and the mechanisms that drive them, and to develop new treatments to treat them. This new research method could be a game changer in psychiatric treatment.
 
The success of the experiment is promising because scientists can build pieces of human brain tissue using stem cells grown in petri dishes. They have already done this with more than a dozen brain regions. But these experiments are challenging.
 
They could aid early testing of drugs
 
Research teams have developed a technique for growing brain organoids from human stem cells. These organoids are tiny models of the human brain and are useful for studying how it works.
 
This technology has many applications for drug testing. For example, scientists have used human brain organoids for testing drugs for psychiatric conditions, such as schizophrenia. These organoids have been successfully implanted and integrated into newborn rats. Researchers are now able to study the effects of drugs on the brain of these animals in a much more accurate way than they could previously.

Disclaimer: Information presented here is for educational purpose and not medical advice. Do not self-prescribe medicines. Consult a registered medical doctor before taking medicines. 

Related Reading:

1. Kaizen for pharmaceutical, medical device and biotech industries​
2. How to cut costs strategically using Kaizen
3. Top 30 Continuous Improvement Tools

Follow Shruti on Twitter, Facebook, YouTube, LinkedIn

Categories:  Life Sciences | Innovation Management

Keywords and Tags:
#stemcells #organoids #organoidsforstudyingmentalhealthdrugs #mentalhealth #antischizophreniadrugs 

​Idea Management is an important and integral part of Innovation Management. There are two schools of thought about idea generation and its management. First, out-of-the box thinking and second, inside-the-box thinking.  

​Out-of-the-box thinking as the name suggests is seamless creativity. While inside-the-box thinking means there is a boundary within which ideas are generated and managed. Wider the boundary, deeper gets the creativity.

And if more ideas have to grow from being on paper to becoming a commercialized product, inside-the-box thinking is the way to go.

​Because, for a business there is a budget to follow, expenses to limit, ROI to gain and investors to take care of. Every penny spent on innovation is accounted for i.e. where profits aren’t earned, the R&D expenditure gets written-off. And the objective is always to get the most out of the buck.

Inside-the-box thinking essentially means, innovation is done with a target product profile in mind. The aim is to generate large number of ideas to maximize number of potential ideas-to-product pathways. The picture above depicts innovation management via inside-the-box idea generation.
​
Once an idea qualifies for becoming a new product candidate, it must successfully pass through the various stages of product development thru obtaining market approval and subsequent launch, to find its crown in the marketplace.

As you will note from the above picture, not all ideas gain ‘new product candidate’ status. So, what happens to the ideas, that do not get selected. Do they get dumped permanently? The answer is No.
​
Let's analyze the ideas further- Usually, ideas get rejected because of three main reasons-

a. Technically idea commercialization is not possible.
b. Profit wise the ideas are not feasible.
c. The company is not keen to go for such a new product at this point of time.
​
​Hence, let's first segregate the rejected ideas into three buckets as per the above classification and then review them one-by-one.

​Ideas that are not technically feasible now can become great candidates later.

​A classic example of this is Sildenafil citrate tablet. This drug candidate was first evaluated as a treatment for chest-pain. Later, the company found an alternate use of the drug candidate and channeled this molecule thru the development pipeline to launch as Viagra- rest is history!
​

​The second lot comprises of those ideas that seem not feasible in terms of profitability. Meaning technically, they may be good, but cannot be commercialized because of high cost.

​One of the classic example for this is- the Microwave oven.

Microwave oven was first invented in the 1940s, around world war II era. The first commercialized microwave oven was huge, weighing 750 pounds and costed $2000 USD. Obviously, it did not sell well.

The company worked on the technical aspects and the first domestic microwave was introduced in 1955. And that did not sell too. After many design changes, microwave oven gained access to our homes in 1967 and by 1975, they were in full demand.

A point to note here is that microwave oven invention was both accidental and EVOP (Evolutionary Operation) product development. Also, note the development timeline, from 1940 to 1955 to 1967 to 1975. i.e. 27 to 35 years from being an 'idea' to 'profitable' commercialization!

It is well-known that EVOP drains money. To avoid such bleed, concept of Innovation Management took shape. I shall be writing more about innovation management in a separate blogpost. For now let us visit the third lot of rejected ideas.
​

Now, let us consider the third lot of ideas- those ideas which the company is not keen to go for at this point of time.

​Such ideas must be worked upon in the lab to develop a small prototype. While prototyping, ensure that you build-in anti-counterfeit measures in your product. Then file a provisional patent application to stake claim to your idea and simultaneously block competition.
​
After filing a provisional patent application, you have 12 months to file final PCT patent specification. Additionally, you get thirty months (from date of priority filing) to enter national phase i.e. do national phase patent applications in the PCT countries of your choice.

Hence, the company gets close to 30 months to scout for buyers who would be interested to license your product. Since a prototype as well as the patent is available (for the invention), you stand to gain better licensing fees at the negotiation table.

Also, resources need not be pumped-in immediately, as product development, process validation, scale-up and other related activities can take place over 12 to 30 months timeframe. This has dual advantages- first, R&D budget can be better managed, and second, R&D infrastructure can be better utilized.

Moreover, provisional and/or final PCT patents being intellectual property can be mortgaged on a short-term basis to banks or other financial institutions, to secure funds which in turn may be pumped-in to conduct research work or use it for other areas of the business.
​
The philosophy of Innovation Management is to ensure, every penny spent on R&D must bring-in good return on investment (ROI) both in terms of profits and company reputation.
​

​There are multiple pathways for conducting idea and innovation management successfully. The content presented above is one such pathway.

Also, the choice of idea and innovation management pathways depend on various factors such as- industry vertical, product mix, operating markets, corporate goals and more. Therefore, selecting the best pathway which will yield faster ROI with a first-mover advantage is crucial. 

Your choice of idea and innovation management pathway and its flawless execution will determine your business's very existence, growth and resilience.

An important question to introspect is- Are you building an idea and innovation management strategy which is the best for your organization?

​To know more about how idea-innovation management can help your company expand its product offerings, increase profit and earnings per share for investors, reach out to Shruti.

​Related Reading:​

1. Disaggregate between being lucky and being good in business: Portfolio Management and Continuous Improvement are the two winning tools in the new normal …
2. 8 Steps to becoming a Customer- centric organization
3. Dos and Don’ts Of Rapid Innovation
4. How to transform businesses via portfolio management- Part 1
5. How to cut costs strategically using Kaizen

Follow Shruti on Twitter, Facebook, YouTube, LinkedIn

Categories:  Innovation Management | Operational Excellence

Keywords and Tags:
​​#innovationmanagement #patent #ideamanagement #stopcounterfeits  #blockcompetition

Lean Innovation, Hoshin Kanrii and Six Sigma (DFSS- design for six sigma) for product design, development and manufacture via 3D Printing (i.e. Additive Manufacturing) Case Study by Dr. Shruti Bhat, gives benefits of applying Hoshin Kanrii, DFSS and Lean Innovation techniques to a drug product design, development and manufacture via 3D Printing.

The study findings however can be extended to product development across other industry verticals.

Some of the benefits of applying continuous improvement methodologies such as Lean, Hoshin Kanrii and Six Sigma to research and product development include-
​

1. Application of Hoshin Kanrii and DFSS along with Quality-by-Design (QbD) to product development usually brings novelty in product features, process characteristics and/or product utility, thus favors patenting and generation of additional revenues.
2. While, Lean Innovation creates products with effective and efficient production and packaging processes, with no scope for ‘rework’ thus, better Operations Management.

​The study uses Isosorbide dinitrate (ISDN) as model drug, although other drugs can also be used in this therapeutic platform.

2x3 Factorial Design of Experimentation (DOE) was used to design drug beads. The results were evaluated for regression analysis, Main and interaction effects of parameters, Contrast Column Dispersion effects, Normal Probability Plots, Pareto and Process Capability Analysis.

The drug beads were further used as ‘starter’ seeds to develop MUPS (multi-unit particulate system) anti-angina ‘rate-programmed’ drug product.

The pharmacokinetic- pharmacodynamic (PK-PD) modeling was done using Doebrinska- Welling equation. MUPS were developed using 3D Printing technology via a 12- Run Plackett Burman DOE.

All hardware, software applications involved in the study were pre-validated. All qualifications i.e. Design Qualification (DQ), Installation Qualification (IQ),  Operational Qualification (OQ), Performance Qualification (PQ) for each experimental run was done as per USFDA guidelines. Data was analysed statistically using Paired ‘t’ test. Reliability Testing was conducted as per ICH current Good Clinical Practices (cGCP) guidelines.

Product prototype and Reference listed drug (RLD) were compared for bioequivalence compliance. The prototype is ready for commercialization.

This study also gives benefits of using 3D Printing technique for pharmaceuticals and how the technique can be applied to product development across other industry verticals!
​
For details on how to apply Lean Innovation, Hoshin Kanrii and DFSS to product design, development and manufacture, watch the video …

To learn more about how Shruti can help your organization achieve new heights or to book a Workshop, Contact Dr. Shruti Bhat via Form or WhatsApp

1. Workshops & Online Courses for Pharmaceutical Research and Drug Product Development by Dr. Shruti Bhat
2. Kaizen for Cost-Savings and Organizational Excellence
3. Mistake-Proofing Pharmaceutical Products: What can we learn from Valsartan, Losartan and Irbesartan recalls?
4. How to develop a culture of ‘Continuous Improvement’ in an organization?
5. ​Innovation is not an initiative. It’s a business process !​

​Follow Shruti on Twitter, Facebook, YouTube, LinkedIn


​​Keywords and Tags:
#ContinuousImprovement   #Innovation  #Lean  #HoshinKanri  #DFSS  #DesignForSixSigma

To learn more about how Shruti can help your organization achieve new heights or to book a Workshop, Contact Dr. Shruti Bhat via Form or WhatsApp

Follow Shruti on Twitter, Facebook, YouTube, LinkedIn


​​Keywords and Tags:
#ContinuousImprovement #Innovation #BusinessTurnaround #Kaizen #JustinTime #HoshinKanrii #QbD #DOE  #Lean

To learn more about how Shruti can help your organization achieve new heights or to book a Workshop, Contact ​Dr. Shruti Bhat via Form or WhatsApp

Follow Shruti on Twitter, Facebook, YouTube, LinkedIn


​Keywords and Tags:
​
#ContinuousImprovement   #InnovationManagement  #QbD #qualitybydesign #DesignThinking  #DOE  #Kaizen #JustinTime  #LifeCycleManagement

To learn more about how Shruti can help your organization achieve new heights or to book a Workshop, Contact Dr. Shruti Bhat via Form or WhatsApp
​
​​Follow Shruti on Twitter, Facebook, YouTube, LinkedIn


​​Keywords and Tags:
#ContinuousImprovement   #Innovation