Group CTO of Techniche overseeing the technology strategy for the Statseeker and Urgent lines of global business.
Getty
For decades humans have been cultivating the concept that “teams” are more powerful and effective than the sole individual. Professor Leigh Thompson of the Kellogg School of Management at Northwestern University defined a team as “a group of people who are interdependent with respect to information, resources, knowledge, and skills and who seek to combine their efforts to achieve a common goal.” Robert E. Cole of the Haas School of Business at the University of California, Berkeley studied the evolution of the team concept in industrialized countries from the 1960s to the 1980s. Whether we call them "autonomous workgroups" or just "teams," the result is often higher productivity with more superb quality. From my experience managing engineering teams, my preferred high-tech strategy has been to form "small teams that move fast." Enable a small team of A players and turn them loose on a mission. Amazing things will happen. Let's fast forward to today and even a bit into the future where artificial intelligence enters the discussion. Now we have the concept of "human in the loop" and a new definition of a team that involves humans partnering with machines (AI bots).
The idea of human-computer interaction (HCI) is a topic of discussion across the industry. Some of my favorite articles were from the Stanford Institute for Human-Centered Artificial Intelligence (HAI). Ge Wang, a Stanford professor, discusses this topic as a software design problem: "In turn, we’ve broadened the question of 'how do we build a smarter system?' to 'how do we incorporate useful, meaningful human interaction into the system?'" He describes the benefits of a "human in the loop" as:
1. Significant gains in transparency
2. Incorporating human judgment in effective ways
3. Shifting pressure away from building a "perfect" algorithm
4. Often enabling more powerful systems, not less
In my recent focus on the use of AI in the fuel retail and convenience store industry (a market my company serves), I've noticed that the subject of AI as it relates to automation is a hot topic. However, the same AI and automation discussions are going across many vertical markets today. The main idea behind the design of AI-based automation systems is that you start with a "problem" to solve. The goal of the solution is that it pays off as a high-ROI investment. Many vendors may try to attack this problem by building "smarter automated systems" to remove human on-site retail staff members or fuel operator employees from the picture altogether.
Another example is that AI image recognition may be able to improve safety on what they call the "forecourt" (the area of the gas station relating to fueling your car). Perhaps AI could deliver an automated alarm when someone drives off with the nozzle still in the car. Other vendors have a vision that through natural language processing (NLP: think Alexa), you will speak to future fuel pumps and interact with them without a human staff member participating in the transaction. BP, for example, tested an AI-driven gas pump in 2016, and you can now use Alexa to pay for gas. This AI-generated interaction might deliver an elevated and personalized "customer experience" where you can fuel your car and maybe order a latte and a snack without any human involvement. Also, the AI system could warn you of safety issues or guide you toward safer behaviors. The AI systems could even detect fraudulent activity. This type of automated self-serve experience without a retail staff member in the loop might also be advantageous during the pandemic.
Because of AI, machines now have "vision," and they can talk to you. However, in software design and current-day computer science, we no longer build monolithic applications but focus on cloud-native architectures that include data pipelines connected to distributed microservices. These microservices are relatively small applications that collectively work together to provide the overall functionality of a system. Each one may have a different function or expertise, a data source related to that function, and an API that connects it back into the overall system. I think of these microservices as AI bots. As I have written about on Forbes previously, they can swarm. Swarming AI bots can collaborate, cooperate and reach consensus. To get back to the concept of teams and the inclusion of HCI in the system's overall design, AI bot specialists could automate a function accurately every time without fatigue. This assumes the AI bots have a healthy diet of useful, clean data. They could make great team members in future AI-based automation systems. However, AI is not perfect and can generate false-positive errors unless the bots are tuned and maintained. Having a human in the loop could enable the addition of classifiers in supervised ML models similar to teaching a child how to learn. A simple example may be to talk to the software and tell the system "yes, you are right," "no, you got it wrong" or "I don't know if you are right or wrong." A human feedback loop can always help drive accuracy in AI machine learning models.
For the evolution of AI-based systems in the fuel retail industry, we need to be extra careful when we're handling fuel. This is why the fuel retail industry is heavily regulated: to keep consumers safe and for environmental compliance reasons. As we see new hydrogen-cell automobiles emerge, the electrical charging stations will also need to be safe to use and safe for our environment. In the future, teams of human domain experts and their AI-bot sidekicks could make sure critical assets in the fuel retail industry are safe, always available, performing at their optimal state, and protected from cybersecurity threats. Whether they are facility or IT-focused, your operations teams should not design the human experts out of your AI-based automation systems of the future. Instead, incorporate them into the system design as a feedback loop to improve quality and intelligence. The results should exceed your expectations as you roll out these innovations as part of your digital transformation.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
