Chris Nielsen, Founder and CEO of Levatas – Interview Series
Chris Nielsen is the Founder and CEO of Levatas. Levatas builds end-to-end AI solutions, machine learning models, and human-in-the-loop systems that supercharge the way businesses are automating visual inspection.
Levatas and its patented machine learning technology, the Cognitive Inspection Platform™, fully automates industrial inspection programs for its global, market-leading customers. The Cognitive Inspection Platform™ integrates with advanced robotics, cameras and drones, as well as pre-trained or custom built ML inspection models to deliver end-to-end automation for industrial inspection uses cases.
Based in South Florida, Levatas serves both the regional and global marketplace, working with industry leading clients like BMW, AB InBev, Dow Chemical, Boston Dynamics, Praxair, Johnson Controls, NextEra Energy/FPL, Ryder, Royal Caribbean, PGA of America, Carrier, G4S, HSBC, and more.
Could you discuss the genesis story of Levatas and how it originated from you losing your job at a software company?
Back in 2006, I was working in Sales for a software company focused on white-label anti-malware software for large telecommunications customers. While in that role, I developed a process for designing mockups of the software that helped me close more deals. I was doing well but the company itself fell on hard times. Later that same year, I was laid off along with many members of the team.
From that experience, I took my rudimentary digital design skills – and a positive entrepreneurial attitude – and started offering my custom software design and development services to local businesses in South Florida. Then things snowballed. One small business customer referred me to a medium-sized customer, and my new accounts started getting larger and larger. As more referrals poured in from happy customers, I had to begin hiring development professionals to keep pace with the growth of the business and our expansion into digital design. We quickly became a jack-of-all-trades digital agency, building anything from websites to e-commerce platforms, to backend software integrations – even offering digital marketing services.
Levatas at first was initially a general-purpose digital agency, could you discuss how Levatas then transitioned into AI?
While the agency’s general-purpose, kitchen-sink approach was good for growing revenues, we recognized that it would be hard to maintain quality and consistency as the suite of offerings expanded. We decided to narrow our focus; moving away from design and development services and focusing exclusively in the field of artificial intelligence and machine learning solutions.
While it may appear to be a big jump – going from a digital agency that offers consulting services to building an enterprise SaaS solution focused on machine learning – it was actually a natural and organic transition.
We’d been working with some of the world’s largest companies, building custom digital solutions based on their data and behind-the-scenes systems. Across multiple platforms and industries we spotted clear and consistent technology gaps that, to us, seemed like market opportunities. Ultimately, we decided to build solutions and products to fill those gaps and, in 2020, Levatas officially pivoted from professional services and consulting into AI/ML software product development. It was the right move.
What was the key moment when it was decided that Levatas would focus on machine perception by using both natural language processing and computer vision versus being a general-purpose AI company?
As a non-technical founder of an advanced technology firm, I’ve gotten good at listening to the incredibly smart people on the Levatas team. It was my business partner and CTO, Daniel Bruce, who set the vision for Levatas to focus on computer vision solutions. Then he sharpened that vision further into “automating industrial inspection solutions.”
My first thought was that this would be too small of a niche, and that we may not find enough customers to fulfill our business growth goals. I couldn’t have been more wrong. It turned out that this is an entire market unto itself, full of huge global enterprise customers that need exactly what we were building.
What’s more, there’s a rapidly expanding field of advanced data capture hardware manufacturers – namely: robots, drones, cameras, IoT sensors, etc – also looking for the solutions we were building at Levatas. This company pivot happened in two somewhat distinct stages over the last 5-6 years. The first stage saw us move from general digital transformation consultants, to an AI/ML specialization (but still as consultants). The final stage of our evolution saw us move out of professional services into the new software product development business model, which is who we are today.
Levatas has partnered with one of the most exciting companies in the robotics space – Boston Dynamics – Could you share some details regarding this partnership?
Honestly, it’s hard for me to talk about our Boston Dynamics partnership without sounding like a total fanboy. [laughs] That said, getting to work alongside the people and Spot robots from Boston Dynamics has been one of the most personally and professionally fulfilling things I have ever done. My team feels the same way.
Not only are they creating the world’s most advanced and capable dynamic mobile robots, but they are simply great people to work with. The bottom line is that the Spot robots come ‘out of the box’ with market-leading athletic intelligence and physical capabilities. What they still need, however, is “on-the-job training” of sorts, enabling them to understand their environments from a cognitive intelligence standpoint. That’s where Levatas comes in.
Our industrial inspection models and Cognitive Inspection Platform enable the Spot robots to inspect critical elements of our customers’ facilities, allowing them to understand what they are seeing and how to react based on the findings. While the Spot robots are capable of so many things, we typically find ourselves deploying them in safety, security and preventative maintenance use cases. These use cases are not specific to any one industry, but we’re seeing a lot of demand in the Electric Utility, Oil & Gas, and Manufacturing spaces alongside Boston Dynamics.
Why is analog gauge reading such a pain point for manufacturers?
You wouldn’t think of analog gauge-reading as a particularly exciting arena for innovation. But for the professionals who are charged with operating, maintaining, and delivering the outputs of those facilities, it’s a big deal.
A given industrial facility may have thousands of analog gauges that monitor various industrial equipment. Right now, personnel have to constantly monitor those gauges (manually) in order to ensure facility uptime and on-target productivity. While digital gauges are available, many facilities operate using legacy equipment that is designed to last decades. Sensorizing thousands of pieces of machinery can cost multiple millions of dollars. It’s also super costly to have incredibly intelligent and capable humans spend their days, every day, walking around the facility to sight-read and report on these analog gauges. Not only is manual monitoring highly inefficient, it can easily fall behind amid worker shortages and more pressing maintenance responsibilities. And if equipment fails because it isn’t checked as regularly, that can lead to even more costly issues.
In contrast, a mobile robot can walk around the facility on a set schedule, conducting those same inspections autonomously using Levatas software. Deploying a robot introduces next-level consistency, reliability, and accuracy with this type of data capture. It also frees up the human employees to spend their time on higher-value tasks for the business – doing work that can only be done by a human.
How does Levatas solve this problem with autonomous tech?
Simply put: an industrial solution that requires manual human operation offers little-to-no ROI. Our customers will not buy it. That’s why all of our hardware partners provide solutions with full autonomy. Their devices create the inspection routes, run the inspection models, and return to their power sources to recharge – all on a loop.
Human workers will still keep an eye on these automated solutions, making sure they’re working as intended. Much like any junior employee who is in training, AI is not yet accurate enough to make the perfect analysis and decision every time. We design our technology to recognize when it must bring a human into the process to help make the right call. In our field, this is called ‘human in the loop’ workflow, and it’s part of the Levatas platform. Overall, the goal is to consistently decrease the time that humans spend on monitoring-related tasks while still ensuring that human workers are kept apprised and always hold decision-making power.
What are some other use cases for Levatas?
In addition to analog gauge detection and reading, we also provide thermal anomaly detection, person detection, robotic collision avoidance, safety compliance monitoring, and a host of inspection model capabilities based on change detection machine learning. When our customers have needs that are not yet met by our existing “off-the-shelf” inspections models, we have a team that works with the customer to develop custom solutions.
While we are excited about our work deploying the Spot robots, the automated inspections solutions from Levatas are also deployed on drones, camera networks, and can be integrated with any other type of data capture devices – such as industrial IOT sensors.
Could you discuss some of the challenges in launching an AI company without being super technical and not knowing how to code?
I’ve always relied on my team of incredibly smart developers to get the job done, and to guide us down the right paths from a technology strategy standpoint. When it came to actually starting the business, I like to think that I had the right mix of ‘can do’ attitude, a positive outlook, and the entrepreneurial spirit that led me to make the initial jump.
Ever since the moment of that first plunge into the deep end, Levatas has been all about the team, and building this together. In short, thanks to the team I managed to build around me in the early days (and to this day), my personal lack of technical capability hasn’t been a big hurdle as we grew the business.
Is there anything else that you would like to share about Levatas?
We just completed our seed round capital raise earlier this year, effectively filling up the rocket ship with fuel. Our solutions are seeing validation in the market with our amazing enterprise customers, and our pipeline is growing by the day at this point. Some exciting new customer announcements will be coming out in the next few months, and we will be announcing some world-first product features later this year. Stay tuned!
Thank you for the great interview, readers who wish to learn more should visit Levatas.
- NVIDIA: From Chipmaker to Trillion-Dollar AI Powerhouse
- Laura Petrich, PhD Student in Robotics & Machine Learning – Interview Series
- Liquid Neural Networks: Definition, Applications, & Challenges
- Patrick M. Pilarski, Ph.D. Canada CIFAR AI Chair (Amii) – Interview Series
- AI Leaders Warn of ‘Risk of Extinction’