What is human factors strategy? My understanding of human factors strategy comes from years of experience working and building medical devices and combination products as well as from observing many errors and failures and yes, committing them myself. Sure the human factors curricula at schools teach us about psychology of users, ergonomics, setting up and running studies, risks and risk mitigation, etc. but these classes do not have the ability or scope to cover what it means to take a complex product from feasibility to production and show you the successes and failures at each step. The true understanding of utilizing strategies and tactics come from the experience. The question of human factors strategy is best answered with an analogy to avoid getting caught up in generalities or semantics:

One of my pastime passions is climbing mountains and I have been fortunate to have learned from my hiking partners and the mountains, many failures as well as successes. As strange as it may sound I see striking parallels between preparing and executing a major climb and implementation of human factors strategy for a critical product. They both can be done in the right way with proper understanding the objectives, planning and preparation. Also, they both can be done in a very sloppy fashion and end in disasters, as they happen everyday.

In mountaineering  is it critical to identify the peaks you want to climb (set an objective), identify the path to the top, select teammates, practice, prepare for the climb, study weather patterns, select an ideal time and hopefully if all goes well summit as planned. One could expect that planning and preparation is 90% of the actual work for high mountains. I remember when we were climbing Rainier we met on much smaller mountains for months and practiced our hikes with loads that go heavier and heavier with each hike. Then we met at Mt. Shasta (a formidable peak) and practiced pulling our sleds loaded with our gear and spent many days practicing techniques like rope travel, crevasse rescue and study of avalanches, etc. In addition to doing these preparatory climbs that were getting harder and more technical, we selected key teammates, practiced being a team and understanding each others needs, moods and behavior, reading maps, identifying and dealing with hazards and elements, resting, hydrating, acclimatization, and much much more. I have seen many novices who would show up at the base camps and start hiking frantically, only to get tired and discouraged before the main push. Or those who would take off at a fast pace and never turn back to see where the rest of their team was. Some climbers would put their heads down and just go without ever consulting a map. Many would end up getting lost or take a wrong path. The right strategy for a big climb is to build a team  and relationships you can rely on. Each team member will have to put in as much as they expect to take out. Each person must be willing to take required risk to save a team mate in trouble and expect the same from others. This team building requires that everyone has a common goal and is synced in mini-goals as well as the ultimate goal along the way. The team has to commit to a common goal and be willing to do many hikes, practice techniques, live together, and connect with each other (and yes, in some cases put up with a difficult person) before attempting the big one. In planning the team has to study the weather conditions and consider all external factors as well as teams readiness in planning and executing the final climb.

Significant parallels can be drawn when you are implementing human factors engineering for a product. Sure it is easy to set up individual studies on the fly and modify the design based on the study outcome but this is no different than showing up at the base of the mountain without a plan, a map, a team, or any practice and expecting to summit a major mountain. Many times I have seen the teams start modifying design of devices or IFUs after individual studies (or even after the first few participants right in the middle of a study) only to revert back the design in the next study. Qualitative nature of the HF studies and the fact that they are run with smaller number of participants can result in changes that will not work for the end-user in the final design. This type focus on short term activities can result in going back and forth in the design. It can also result in conflicts and frustrations in short term and loss of confidence in human factors capabilities as part of product development. 

To implement the right human factors strategy the HF team must start with users (both internal and external) and understanding their needs. Internal users are the key stakeholders (marketing, product engineers, quality, regulatory, project management, manufacturing, etc.) These teams will define the product that needs to be built and they have critical input to human factors plans. External users are the end users of the product but that is too narrow of a definition.  In reality to understand the comprehesive list of users we need to consider anyone who come in contact with the product. For a medical device the external users could be the product shipping and handling, distributer, installer, operator, patients and healthcare professionals. Once the users and their needs are understood and defined it is important to have a human factors plan and clearly spell out the goals of the human factors activities, roles and responsibilities and number of studies, the expectations for the final design of the product, user interfaces and how different users will be interacting with the product from prototype stage all the way to the final product. HF plan should be a living document that aligns studies objectives and outcomes to the broader product development actives and allows for holistic view of the studies and design changes. For a successful HF Engineering, the HF plan should have a map and timeline and should be reviewed with stakeholders and capture their input and needs. Each team member should be committed to bring in their expertise and knowledge and contribute to the success of the overall plan. In addition to the plan, the HF experts need to make sure the team take the time to reviews the outcome of individual studies in light of overall objectives of the product development and users needs. Taking holistic view of the objectives and data from studies along with consideration of the usability requirements will really ensure the success of the program and avoid in getting caught up in endless circles of design and redesign confusion!  

HF Strategy is about planning for success and meaningful iterative studies before attempting to design (or redesign) a product but it is a lot more than that: HF Strategy is about understanding users and users needs, assembling a team of stakeholders, building relationships, taking holistic view of the study data and separating real issues from the noise. You can always run up a mountain but without proper planning and preparation usually the mountain will have the final say and the summit may remain elusive. Respect the mountain!

Human Factors is understanding of the limitations of humans and applying those understandings to the design of machine interface to make it usable for the average human being. As the machines have got more sophisticated the task of interface design have become more and more important and intense. Whether it's design of web page for salient navigation or the design of an airplane cockpit with all the grouping of the buttons, shapes, color coordination and optimal operation the effort has been to simplify the user interfaces for the humans with limited physical and cognitive capabilities and attention span. Contrast that with the explosive growth in memory capacity, learning capabilities and adaptability of the machines. 

Before transitioning to discussing AI, I want to discuss the free app on my iPhone for playing chess, which I play casually when waiting for my kids or when I want to get depressed! The app is a standard chess app with the typical user interface and the difficulty level is usually set to 4 (out of 10). I am fully familiar with the interface and generally don't make too many mistakes in moving the pieces (no UI design issues). The tiny chip in this phone beats me about 7 out of 10 times. If I set the difficulty level to 9 or 10, I will have no chance of winning at all (I have tried!) This app has no machine learning built into it that I know of. Now let's imagine a scenario where this app could collect all the information from my every move and use it to better it's own gaming strategy and make its moves more sophisticated and unpredictable. Let's further imagine that to make things interesting it could change the shapes and colors of the pieces and the board. I would no chance in hell winning even a single game. 

Why this story? The story points to the limitations of human beings (just me in this case) and unlimited power of AI if it were to be implemented into a chess game with the intent to defeat the human being (think Watson!) It also points to the potential role an AI user interface could play in this brave new world for better or for worse.

User interface design under the current human factors/design thinking paradigm of designing for human limitations may be too small of a box to fit the AI genie in. Smarter UI designs that could adapt to AI needs and not be constrained by the human handicap could be an important aspect of AI/human interaction success. So what is the role of human factors?  Continue designing for human limitations as it's done now (and in the past) and be the drag force in the AI revolution or something else.... - Am I the only one seeing this?! I recently saw a Duke University course on how to build AI around humans. Should we pretend the emperor is fully clothed... More to say about this later.