In the last few years, “Machine learning” has become one of the buzzwords among the experts across various Industries. In simple words, Machine learning is the application of Artificial Intelligence (AI) that enables systems to automatically learn and improve from experiences without being programmed. Machine Learning has diverse applications including Product Recommendations, Online Customer Support, Virtual Personal Assistants among others. AppliedML (Applied Machine Learning) is one such company Using Machine Learning and Artificial Intelligence to Analyse Human Behavior in Industrial Context. We recently sat down with Roy Daya, Founder and CEO of AppliedML to know more about his company and Artificial Intelligence in general.
Tell us more about your journey as an Entrepreneur. How is AppliedML helping executives in the Industrial Context?
I was born an entrepreneur I guess. All my life I searched for my own path in anything that I did. I was always involved both on the business side of launching profitable businesses and on the technology side. Starting a business in the context of cutting-edge technology brings many opportunities and openness in the market for you to wow customers and be accepted as a small and young company because you have learned to do someone that can be worth to your customers a lot of money and they are eager for you to succeed and build success with them. I love this position much more than negotiating sales of a common product or service where there are thousands of competing providers and customers are not excited about your offerings.
In the last 15 years, I have launched and managed two successful businesses but was involved with hundreds of startups along the way assisting them with their technology, innovation or with finding the right place for them in their chosen business ecosystem. I love challenges and am always ready to help fellow entrepreneurs internationally. The first company I launched was Digital-Clay back in 2003 and I managed it until 2012 when I realized that I spent most of my time exploring advanced analytics and ways to help businesses understand the story behind their data.
In 2012 after I left Digital-Clay (which is still doing great even today) I launched AppliedML as my specialized consulting and technology company where I provide AI consulting to startups and larger businesses internationally. As we grew, we started to develop our own products. Starting the business around a service allowed me to start a profitable business and not rush into choosing my products and crafting them to fit specific business needs.
Our products ended up being in the field of computer vision where a computer can take an image or a video clip from any security camera and accurately analyze the scene including behavioral aspects of people and animals and their interaction with various objects. This allows us to provide a solution that detects when something is going on which needs to be addressed in some way and generate the right alert to the right person at the right time when it is still possible to take action and minimize the cost or risk or to take advantage of the situation in some way. For example, a person walking near a container with chemicals and starting to show signs of distress or objects being placed on a top shelf in a way that can fall on someone’s head. There are many applications and our customers help us extend our perception of what can be done with each new project.
AppliedML is a great opportunity for me to travel the world and help solve operations and safety-related problems that are very difficult or expensive to solve in other ways. AppliedML exists as a service provider since 2012 but our products are only out for the last year so it really feels like a young Startup now as we develop the company as a product company and not just as a service provider.
Though we are witnessing a dramatic surge in the application of AI across various industries, “AI is still viewed as an exotic, cutting-edge technology shrouded in mystery”. What are your views on this?
I think that AI is a buzzword that became very popular lately but the concepts are not new. What is new, is much greater availability of computing power cheaply and the availability of amazing tools and technologies a developer can use to create solutions fast.
Statistic analysis is used to analyze data and better understand what happened in the past. AI takes historical data and tries to identify generic rules that can be used when new data arrives to predict the outcome. AI is the proactive side of statistics. Statistics will segment your customers into groups for example and AI can tell you if the next lead that just entered your online store will buy anything and if so what and at what price and for what reason.
I think now the technologies are ready and customers are getting more and more aware of the potential but still, there is a great need for more experts with experience that will create a successful solution using AI technologies. I know companies that are willing to pay huge sums of money but simply cannot find the right people.
How does your company use Artificial Intelligence to Analyse Human Behavior across various industries?
We mostly focus at AppliedML on industrial and manufacturing customers simply because their problems are so much more complex than a services business or other type of business. There are immediate physical risks and hygiene issues which may have grave outcomes if not immediately addressed.
We sit with each customer and take the time to understand the scenarios that we need to analyse so not only we can deeply understand exactly what is going on from each camera view but we also are able to alert when something happens that we are not supposed to alert from but because it is so much different than what is expected we can alert that something strange is going on and so address even unknown situations.
We come to our customers with a rich library of over 50 industrial interactions and mini solutions that we can tailor and combine to deliver in a few short days. A solution that is very unique in the market. We connect our software to any existing camera feed and many types of sensors to understand the current scenario using our understanding of what interactions are expected and what order of actions is expected.
When we find there is an issue we can sound a siren or shine a light as a notice or even send a message to someone at another location, for example, to get help. We can also send notes about locations of objects, their utilization history, and any observed damage or malfunction.
Internet of Things devices are big nowadays. How do AI and IOT go together?
IOT is a way to allow machines, or really any object to send a computer server details about itself. It can be its location and other parameters. Machines can send operational data, a car can send location and efficiency parameters for example and a medical device can send sensor data from various sensors it has.
AI is used to take all this data and learn about the relationship between various numbers and outcomes. AI software can be on the server or even on the device itself and create a predicted outcome such as a health condition suspected or car maintenance prediction or other assessment of the IOT device status. AI can also control the IOT device generating the needed corrections for the IOT device to implement to improve its status.
These devices are becoming smaller and energy efficient, and internet data storage is becoming very cheap making the advantages of these systems realistic to many more businesses than before.
How Can AI be used to predict outdoor machine behaviors?
AI excels in places where the operating conditions are not constant. When nothing changes you don’t need to predict what will happen as it is a reliable process.
Outdoors, there are extreme or unreliable conditions which make the machines operate differently than expected in normal conditions. Different is expected but you need AI to assess when different is good and when there is a problem. It means you need to predict which behavior is expected and always compare the prediction to what is observed. People that work with or walk around these machines need to be tracked to make sure are not in direct risk because of the weather or the machines around them.
What are some of the ways in which organizations can benefit from the services provided by Applied ML?
Managers come to us to improve safety and operations in most cases. They have a process which works well most of the time but when it fails, it very quickly becomes very expensive and sometimes complex to recover back to normal. If we can let them know things are about to turn into a crisis we can save lives and a lot of time and money.
If you could take an expert that knows everything your organization knows and replicate and put him or her in every room so they can watch over anything that happens in your organization you could save yourself a lot of crises to deal with. This is obviously not possible but today a computer can do this. Also, a computer does not really look at people and hurt their privacy. It is just a sensor that generates an alert when a prediction is made about a situation. Nothing is recorded or saved from that camera other than the alert that was generated.
Artificial intelligence is progressing very rapidly. Digital assistants like Apple’s Siri and Google Assistant are getting better at helping people with routine tasks. However, there are controversies around the use of AI in the development of autonomous weapons. Do you think that the fear around some of the usage of AI is justified?
I think that AI represents a field with very powerful tools. Like any tool, it can be used and abused. It can be abused because of ill intent or because of lack of proper understanding of the limitations of the technology. Not understanding the limits will allow the AI technologies to make a decision in places where the quality of these predictions is very low.
So yes, there is room for concern from overly eager organizations to deploy systems that the quality of their automated decision making is not properly trained as to do it properly may take a lot of resources.
As an AI expert, how do you see AI impacting our society in the coming few years?
AI will be a blessing as we will be able to always get the exact service or assistance or product that is tailored to us. A long time ago everything was handmade and custom. It was a great experience but the time involved in making those products along with the cost of each product was very high making them inaccessible to many people. Later, as we started mass production we could lower product price at the cost of variance. We had to conform to the available models and sizes. With AI it is possible to offer mass customization to many products and services at a low cost as a computer is the one that learns about our needs and makes all the needed adjustments to fit our exact wants and needs.
We know that you are an Avid Traveler. Before we wind up, Would you like to share some tips with other executives traveling to exotic places? Also, what are some of the places you visited in the last year or two?
People are always looking for being connected all the time. They spend the few hours they have free chasing wifi hotspots. I am sure most executives work hard enough. They should learn to take an hour or two and take a step back, go sit in front of the ocean or next to a snow covered mountain and just be. Recently, I was in South America a few times, in Costa Rica and Cancun, I was in Iceland and a few times in France and the UK. I was a few times in the US and have planned trips coming to Russia and Singapore soon. I grew up in Singapore so looking forward to visiting there again.