Joe Lui Of Accenture Industry X On The Future Of Robotics Over the Next Few Years

With the shortage of labor, companies are now looking at how robots can replace some of the lost labor force. See here for example. The truth is that this is not really a novel idea, as companies like Amazon have been using robots for a while now. What can we expect to see in the robotics industry over the next few years? How will robots be used? What kinds of robots are being produced? To what extent can robots help address the shortage of labor? Which jobs can robots replace, and which jobs need humans? In our series called “The Future Of Robotics Over The Next Few Years” we are talking to leaders of Robotics companies, AI companies, and Hi-Tech Manufacturing companies who can address these questions and share insights from their experience. As a part of this series, I had the pleasure of interviewing Joe Lui, Managing Director, Global Robotics Practice Lead, Industry X, Accenture.

Joseph (Joe) Lui is Managing Director at Accenture Industry X leading the global Robotics practice. He has made a career of identifying emerging technologies, evaluating their readiness and opportunity for market, then commercializing them to profitability. Prior to Accenture, Joe was the Vice President and General Manager of Robotics, Computer Vision and Artificial Intelligence at Honeywell, where he founded the Honeywell Robotics business unit and developed the company’s Smart Robotics system solution offering for retail, e-commerce, manufacturing, logistics and transportation customers. He was also previously Director of Automation and Robotics at Amazon, where he led the effort of developing the next-generation autonomous fulfillment and transportation network. At Caterpillar, Joe was Head of the Digital Services and Autonomous System Division, where he led the commercialization of the world’s the largest autonomous fleet for mining applications.

Thank you so much for joining us in this interview series! Before we dive in, our readers would love to “get to know you” a bit better. Can you tell us a bit about your ‘backstory’ and how you got started in robotics?

I am Managing Director at Accenture Industry X, where I lead the Global Robotics Practice. I have made a career of identifying emerging technologies, evaluating their readiness and market opportunity, then commercializing them to profitability.

Prior to Accenture, I was the Vice President and General Manager of Robotics, Computer Vision and Artificial Intelligence at Honeywell, where I founded the Honeywell Robotics business unit

and developed the company’s Smart Robotics system solution offering for retail, e-commerce, manufacturing, logistics and transportation customers. I was also previously Director of Automation and Robotics at Amazon, where I led the effort of developing the next-generation autonomous fulfillment network. At Caterpillar, I was the Vice President and General Manager of the Digital Services and Autonomous System Division, where I led the commercialization of the world’s the largest autonomous fleet for mining applications.

Can you share the most interesting story that happened to you since you began your career?

The following happened a while back when my then-team and I were automating a brand-new automated distribution center for a client. It’s a good example of how organization and culture need to be an integral part of any automation initiative, and that technology alone will only solve part of your problem.

After we had commissioned all the robotic workcells, we put the entire automated operation in service. However, the operation only stayed up at full speed for one shift. We had to slow down the operation to readjust the workflow. Workers were so busy with exception management, resolving bottlenecks, loading materials, etc., and ultimately, we were only able to achieve less than 40% productivity of a manual operation in that shift.

There are three components that make a successful automation deployment: 1. process, 2. people, 3. technology. Automating a bad process will only compound your operational problems. What is more, you need a process and workflow that is optimized for robotics, which is drastically different from a human-centric workflow.

Can you please give us your favorite “Life Lesson Quote”? Can you share how that was relevant to you in your life?

“Creativity is intelligence having fun!” said Albert Einstein. Ultimately, you need to get yourself out of your daily routine, then creativity will flow. It’s a beautiful thing!

Let’s now shift to the main focus of our interview. Can you tell our readers about the most interesting projects you are working on now?

My most interesting project is to design and implement a fully autonomous warehouse operation. You can imagine it involves a lot of steps and design components: 1. Designing the process that is optimized for robotic workcells, which is drastically different from a human-centric process design; 2. Developing the robotic workcells by customizing off-the-shelf robotic solutions that meet our specific requirements; 3. Designing the artificial intelligence (AI) orchestration engine to interface with various IT and operations technology systems to automate decision making.

The AI orchestration engine design is the most exciting and challenging task. Ultimately, it is the key enabling technology to make any operations fully autonomous. The orchestration engine uses multiple real-time machine learning models to automate decision-making on the shop floor based on historical patterns and knowledge.

How do you think this might change the world?

AI process orchestration, as a key enabler of any fully autonomous operations, can make many people’s jobs easier in so many ways. It also has a key role to play in solving labor shortage.

Consider that today, even managers in automated buildings are required to make lots of real-time decisions, based on their experience, without taking into consideration all available data. This is what I call “localized optimization.”

Nowadays, with the proliferation of industrial internet of things sensors, gigabits of raw data are being generated on the shop floor. Industry 4.0 technology helps us turn this data into actionable insights. This is a good start! However, all these “insights” are being presented to managers on multiple screens in a control room, who are then supposed to process them instantly and make real-time decisions. But the human brain can only process so much data simultaneously. Not to mention these managers need to make split-second calls on the fly. They all try their best to make some decisions (right, wrong or different) with their experience and with whatever insights they can process at that moment.

Moreover, there is no way we can validate the decisions being made. This is where AI can help. Computers can manage gigabytes of data in a second. Machine learning can detect patterns from the data lake in a relatively short period of time. Then, AI process orchestration — essentially an intelligent scheduler — dynamically allocates resources and assigns tasks to individual workcells aiming to optimize the overall efficiency of the building based on simulation and historical data from various sources. This kind of improved building and network efficiency will reduce dependency of human labor and provide opportunities for workforce upskilling.

Keeping “Black Mirror” in mind, can you see any potential drawbacks about this technology that people should think more deeply about?

AI is moving at a very fast pace. Organizations that apply AI to make their operations fully autonomous need to build trust with the public and with the operators. They should be held accountable to their customers and employees. Designing, building and deploying AI-powered systems must follow certain principles, which I would like to highlight here:

Transparency and Explainability
As AI technologies become increasingly responsible for making decisions, we need to be able to understand how AI systems arrive at a given conclusion, taking these decisions out of the “black box.”

Human has Ultimate Control
Machines don’t have minds of their own, but they do make mistakes. We should have risk frameworks and contingency plans in place in the event of a problem. It needs to be crystal clear who is accountable for the decisions made by AI systems and how we fix the problems if and when they occur.

Safety

As we begin to embed more AI features into automation solutions, we must consider the impact that robots and AI technologies have on humans and society with regards to physical safety and beyond. We need to verify that an AI-based product incorporates failsafe mechanisms, which allow operators to take over control when automated AI applications reach the limit of their competency.

Data Security
Companies must design privacy, transparency and security into their AI programs from the outset and make sure data is collected, used, managed and stored safely and responsibly.

What are the three things that most excite you about the robotics industry? Why?

• The number of robots being installed is increasing every year, even during COVID.
• AI has moved from being science fiction to becoming an enabling technology for mainstream robotics.
• There are countless innovative ideas coming out on a daily basis. One of my favorites is how breakthrough technologies enable robotic manipulation of soft and non-rigid objects.

What are the three things that concern you about the robotics industry? Why?

• More and more companies invest heavily in deploying robots, but few are able to achieve significant returns because implementation takes too long, and deployment cost is higher than expected.
• Many are getting stuck in the pilot phase and unable to scale their point solutions appropriately.
• There is a lack of innovation in customizing robotic solutions to meet end-user requirements as many system integrators aren’t going beyond integrating off-the-shelf solutions from original equipment manufacturers (OEM).

As you know, there is an ongoing debate between prominent scientists, (personified as a debate between Elon Musk and Mark Zuckerberg,) about whether advanced AI has the potential to pose a danger to humanity in the future. What is your position about this?

Looking at AI that enables autonomous operations, I strongly believe that as long as human has the ultimate control, AI is not a threat. But of course, AI’s direct impact on people’s lives has raised considerable questions around AI ethics, data governance, trust and legality. As organizations start scaling up their use of AI to capture business benefits, they must embrace an approach we call Responsible AI. It’s the practice of designing, developing, and deploying AI with good intention to empower employees and businesses, and fairly impact customers and society. It includes the principles I outlined in my answer to question 6.

In today’s environment, hackers break into the software running the robotics, for ransomware, to damage brands or for other malicious purposes. Based on your experience, what should manufacturing companies do to uncover vulnerabilities in the development process to safeguard their robotics?

As we begin to embed more AI features into robots, we must take cybersecurity extremely seriously. The expanding use of AI controls will demand evolution of safety mechanisms to safeguard AI systems. Penetration tests and cybersecurity certifications must be part of the commercialization requirement before rolling out any AI features.

In my mind, there are a few key concerns that must be addressed in order to “certify” that an AI system is safe. Testing should assess the extent to which AI-driven products behave in consistent and predictable patterns, free from unintended side effects, even when they are required to adapt dynamically to their operating environment. Again, it is ok for AI to fail. But, if it does, we need to make sure AI fails gracefully. Operators need to be informed and ready to take over control when the environmental data departs significantly from the circumstances for which they were trained.

We also need to have mechanisms in place to detect and limit any irrational instructions being issued by the AI algorithms and allow users to take over control when automated AI applications reach the limits of their competency.

Testing should be able to certify that there are mechanisms in place so that the user can rescind AI-driven decision agency in circumstances where the uncertainty is too great to justify autonomous actions. Finally, vulnerabilities in AI networks can expose companies and society to considerable risk if they are discovered or hacked by malicious threat actors. AI systems need to have defense mechanism in place, so when they detect any cyberattack, there will already be a failsafe mechanism in place that can be activated.

Given the cost and resources that it takes to develop robotics, how do you safeguard your intellectual property during development and also once the robot is deployed in industry?

The normal route is to file patents. That said, most of the innovation is embedded in the software, the way the robotic solutions are implemented, and the processes that are optimized for robotic deployment. These are things that are hard to copy or reverse-engineer.

Here is the main question of our interview. What are your “5 Things You Need To Create A Highly Successful Career In The Robotics Industry?

• You need to be curious. You must want to dive deep into the problems you are trying to solve before starting to design robotic solutions. Remember, robotics may be able solve your operational problems effectively, but there are other tools you can use, too.

I once visited an automated plant. The general manager was so proud to tell me they had just allocated 10,000 square feet to deploy AMRs. I asked him why he didn’t just install a cross belt sorter. It could have been cheaper and more efficient. I got no response.

• You need to have a broad knowledge base and a good understanding of some enabling technology: Machine Learning, Computer Vision, Digital Twin/Simulation, Sensors, Mechatronics.
• You need a strong network, inclusive of industry leaders and academia.
• You need extensive hands-on experience — deployment, implementation, customization — and a deep understanding of various manufacturing processes in different industries, warehouse operations and distribution center operations.
• You need to have a deep understanding of business models, pain points and operational challenges in different industries. For example, automation requirements for electric vehicles manufacturing are very different from e-commerce fulfillment operations.

As you know, there are not that many women in this industry. Can you advise what is needed to engage more women in the robotics industry?

I have tried very hard to address this issue for many years. There are a few things we could do:

• Make sure we have a strong candidate pipeline. We need to get more women interested in majoring in robotics, mechanical engineering or computer science in college. I am currently on the Advisory Board of the University of Washington Mechanical Engineering Department, where I have made it a priority to organize seminars that bring female executives from the robotics industry to campus and career fairs as keynote speakers to share their career stories.
• Increase the number of female faculty within engineering and computer science.
• Working with HR teams to actively address the intrinsic biases in the workplace.
• Provide female scholarship in computer science and engineering.

If you could inspire a movement that would bring the most amount of good to the most amount of people, what would that be? You never know what your idea can trigger. :-)

Robots eliminate dangerous jobs for humans and AI reduces stress! The technology gives companies an opportunity to upskill their workforce. Robots can improve our quality of life and make the world a better place. My message to readers would be: “Be bold and accept robotics as part of our lives, as it will continue to improve everything from our jobs to how we will travel.”

Thank you so much for the time you spent doing this interview. This was very inspirational, and we wish you continued success.

About The Interviewer: David Leichner is a veteran of the Israeli high-tech industry with significant experience in the areas of cyber and security, enterprise software and communications. At Cybellum, a leading provider of Product Security Lifecycle Management, David is responsible for creating and executing the marketing strategy and managing the global marketing team that forms the foundation for Cybellum’s product and market penetration. Prior to Cybellum, David was CMO at SQream and VP Sales and Marketing at endpoint protection vendor, Cynet. David is the Chairman of the Friends of Israel and Member of the Board of Trustees of the Jerusalem Technology College. He holds a BA in Information Systems Management and an MBA in International Business from the City University of New York.