AiThority Interview with Ashwini Choudhary, Co-Founder at Recogni
Could you tell us about your interaction with smart technologies like AI, IoT and Robotics?
Automobiles are IoT devices with increasingly more sensors and cameras. Using advanced algorithms, we are able to process data using Artificial Intelligence to enable autonomous vehicles to perceive hazards and identify things such as traffic lights surrounding the vehicle to intelligently make decisions about how fast the car should be going, if it should slow or come to a complete stop, when it can safely change lanes or make a right turn.
While companies like Amazon are now considered logistics companies, there is still a cost obstacle for the last mile delivery. Currently, drivers are hired to take packages from a regional location to each home and business. Robotics such as drones are being tested and adopted for the last mile delivery. Our solutions can make sense of data from cameras and we offer an attractive and necessary technology for last-mile deliveries.
Could you tell us about your journey into tech and how you started at Recogni?
I have a long history of working in Silicon Valley. Prior to Recogni, I was Co-Founder of a company called mPerpetuo, which developed a camera to migrate smartphone UX onto a DSLR. During that time I was in a motorcycle accident caused by a distracted driver in San Jose that required many surgeries and lots of physical therapy. It was there that I began to ask the question, “How do I make sure nothing like this ever happens to anyone ever again?” Eventually, the question evolved into working on this great technology that we believe overcomes a major obstacle for the autonomous vehicle industry. My experience at mPerpetuo allowed me to transition to Recogni because imaging is a very important aspect of perception processing technology.
What is the state of “Autonomous Vehicle’ technology? How do you prepare for this complex industry?
There are five levels of autonomy. Right now all we have achieved is a level of safety also known as Advanced Driver Assistance System or ADAS, sometimes also known as Level 2 Autonomy. Next is Level 3 which requires the driver to intervene within some limited timeframe. What the industry is working on is achieving levels 4 and 5, which requires no driver intervention. A power-efficient perception system design is a vital next step but very few companies can solve the problem. Several startups are working on this and it will be several more years before it reaches the price point to get robotaxis en masse on the road.
What is the Future of AIOps, Computer Vision, and Machine Learning in automobile industry? How do you empower your customers adopt these technologies at a better pace?
In order to keep drivers and passengers safe in autonomous vehicles Computer Vision or what we call perception processing is a huge obstacle that must be solved first. AIOps will enable the technology powering autonomous vehicles to automatically spot and react to issues in real-time such as mechanical failure that autonomous cars cannot resolve on its own. Customers will start seeing the same technologies we are putting into vehicles being adopted in other industries for lots of different use cases and as it scales, the cost will also come down. As their familiarity increases and cost decreases, it will help accelerate consumer consumption.
Take Healthcare, for example, the way an Oncology doctor treats patients is two to three years behind the available research. Using Artificial Intelligence and Machine Learning, we’ll be able to diagnose cancer more quickly with more certainty leading to better prevention, care and quality of life. This will also have a circular benefit to the automobile industry. From the Aviation industry, we can better understand how cabin pressure, air quality, seat programming, and music impact passengers. Then the automobile industry can apply that to understand how to keep passengers more comfortable and refreshed while sitting in a car for extended periods. The same thing is happening with cameras and how we are applying what we learned into perception processing.
Which recent events in Robotics and intelligent driving forced you to relook into the technology offered to the customers?
We closely monitor safety concerns from consumers with self-driving cars. There has been advancements from companies like Tesla and Waymo. One thing that is common in San Jose is how drivers go around Waymo self-driving cars that are programed to be extra cautious and sometimes waits way too long to make right turns, especially on red lights. Another interesting problem that recently came up is the ability to charge vehicles like Tesla during a power outage – like the most recent shut off happening now in California as the local electricity company PG&E shut off power during a high fire danger advisory period. We’ll continue to monitor these types of activities to better understand problems that weren’t necessarily top of mind when we began Recogni.
How do you see AI and ML becoming the core of every business in the next 2-3 years?
Artificial Intelligence and Machine Learning will become more mature but there’s no doubt the benefits are self-explanatory, especially when it comes to Sales and Marketing. For every business, there’s an opportunity to better serve a customer, a cost that can be cut, an efficiency that can be gained and a revenue opportunity. Improving customer experience leads to better brand loyalty, and we all know that reoccurring revenue is less costly than new customer acquisition.
What we have right now are more sophisticated and capital-rich industries leveraging these capabilities to grow and advance their companies, then it will have a trickle-down effect. You will also see technologies such as speech recognition becoming more advanced to understand every accent and dialect, even my funny accent. Then it will be beneficial to hearing aid technology and so forth. As these technologies mature, we will stop talking about AI and Machine Learning as this cool futuristic advanced technology because it will be omnipresent and hiding in plain sight.
What do you think about the ‘Weaponization of Data and AI” tools? How do you create awareness in the society about Ethical AI applications?
With everything, there’s an opportunity to use it for good and evil. We can’t rely on legislation to fix problems like these. We continue to see the problem evolve in cybersecurity as we started with blocking threats with antivirus and now we’re starting to use Artificial Intelligence and Machine Learning to develop more sophisticated ways to safeguard ourselves. But the reality is that as we evolve with the use of software, devices and behavior, the number of threat vectors are going to continue to grow and the amount of data we generate will continue to multiply exponentially. It’s going to take a concerted effort combining Human Intelligence and Technology to effectively thwart problems in the future because legislation lags behind problems we’re encountering in the present.
What kind of training should IT professionals and analysts undergo to prepare for the current market and future industry demands? Do you train your AIOps and ITSM teams with emerging technologies?
As we discussed previously, Artificial Intelligence and Machine Learning are going to be hiding in plain sight. Our society continually builds on what previous generations have taught us and reiterates to make our futures better so working in collaborative environments where AI and Machine Learning is a focus is critical. Getting exposed to it any way you can from taking courses, getting involved in an industry doing something interesting with Artificial Intelligence and Machine Learning or reading about it within research environments will help professionals keep pace and, in some cases, stay ahead of the curve.
Could you tell us more about your Risk Profiling, and how customers benefit from deploying it in their IT stacks?
As we’re analyzing data from cameras and sensors, we don’t use risk profiling.
What is the biggest challenge to Digital Transformation in 2019? How does BigPanda contribute to a successful Digital Transformation/IT Modernization?
Digital Transformation is challenging as it’s a business-centric buzzword that hasn’t been clearly defined and varies from industry to industry. Central to Digital Transformation is taking data from different edge locations and devices and using Artificial Intelligence to intelligently make decisions for the business and its customers. How pervasive it is differs vastly in every industry and segment but over time, the hype will die down and it will become a mainstay embedded into many parts of our lives.
Another big challenge is the skillset of the industry currently but as Artificial Intelligence becomes more pervasive it will become a programming language that anybody can learn similar to how we think about the English language. We’ll always need tools to fix IT issues quicker and intelligent automation, if deployed correctly, will alleviate many of the technical challenges associated with IT.
How should young technology professionals train themselves to work better with Cloud, Automation and AI-based tools?
My son is about to enter college and I’m encouraging him to get as much exposure to Artificial Intelligence as possible taking classes, reading books, getting involved in collaborative efforts and trying things on his own. That’s the only way to learn and grow, roll up your sleeves and dive in.
What AI, ML and SaaS start-ups and labs are you keenly following?
Being in the autonomous vehicle industry I follow a lot of different types of companies but the ones that stand out currently are obviously Tesla both as a business and automobile company and Waymo. Another company that I follow is Mobileye who has been working closely with automobile companies to provide safety-focused Advanced Driver Assist Systems (ADAS) technology that we find in cars today.
What technologies within AI/NLP and Cloud Analytics are you interested in?
Google with its vast resources seems the furthest long both with its NLP and Cloud analytics capabilities. Google can understand my Indian accent and that type of technology if done effectively can become embedded into a lot of different things we use today.
As a tech leader, what industries you think would be fastest to adopting Analytics and AI/ML with smooth efficiency? What are the new emerging markets for these technology markets?
The automotive industry has a huge opportunity with Artificial Intelligence and Machine Learning that will take off rather quickly. We hear about autonomous vehicles but it also touches transportation, delivery robots, and freight deliveries. If you look at Amazon and you see how they have evolved into a logistics company, you can see the type of opportunity there is.
What’s your smartest work related shortcut or productivity hack?
As a leader, the smartest productivity hack is hiring smart people, opening up clear communication channels and utilizing tools that encourage collaboration. This shortens cycles and improves efficiency.
Tag the one person in the industry whose answers to these questions you would love to read:
Being in Silicon Valley and the industry I’m in, Elon Musk is a person whose brain I’d like to pick. He’s been doubted by many and continues to do bold things that prove people wrong.
Thank you, Ashwini! That was fun and hope to see you back on AiThority soon.
Ashwini is a Serial Entrepreneur/Company Builder with >20 years of experience. He built six startups with four exits via acquisition to various public companies. He his highly skilled at identifying gaps in the market, defining the product, and raising venture capital.
Ashwini is extremely capable of building high-performance teams, driving product management and delivering product to the market for first revenue.
The automobile industry has arrived at a crossroads. The transition to electric vehicles (EV) and the vitalized development of fully-autonomous vehicles (AV) has placed a big burden on fitting extraordinary amounts of computational power for artificial intelligence within the energy budget of batteries without affecting range. While battery technology is improving slowly, advances in compute efficiency have stalled as mere Moore’s Law scaling of computational architectures from the past is nearly at an end.
Recogni, with its unique approach to designing a vision-oriented inference artificial intelligence system from the ground up as a holistic module, will deliver unprecedented inference performance at more than 500x better power efficiency compared to other solutions, enabling novel edge processing at multiple points on vehicles to naturally offload central processing needs.
Building on a strong foundation of entrepreneurial team-building and operations experiences, our unique blend of proven track records in high-performance computing and distributed systems, artificial intelligence and machine learning, and imaging and vision systems is empowering us to accelerate the realization of fully-autonomous vehicles.