Work-In-Progress Video 001: Supporting the Development & Evaluation of Human-Machine Teaming Systems
It seems like we’re always working on something we’re excited about. But, once the capability exists and we’ve moved on to refinement mode, we’ve already set our sights on the next thing. That’s the problem with goals—you focus on them while they seem difficult or impossible to achieve, but as soon as you’ve done the hard work to make them attainable, you move on to the next challenge. Unfortunately, this often occurs without even noticing that you’ve accomplished something you weren’t even sure you could achieve when you started.
That’s just human nature. We all do it. But, for SantosHuman Inc., I need to do better. I need to make sure we highlight the achievements along the way. With that in mind, we’re going to start letting you in on the things we’re working on—while we’re all still excited about working on them.
This first installment of the Work-In-Progress series demonstrates a capability goal we’ve been working towards for a very long time. The ability to use Santos technologies in the development and evaluation of human-machine teaming (collaboration robotics, human-agent teaming systems, etc.) will very soon be available within our flagship product, Santos® Pro.
Take a look and let us know what you think. We’d love to hear from you.
The predictive nature of Santos technology means we don’t require motion capture to evaluate human systems integration requirements. However, many of our clients had expressed a desire to use motion capture, and we initially provided this capability in Santos Pro in 2013. While the capability was refined in 2014 to include the fingers, we noticed we had never provided a video tutorial.
In this new video tutorial, we’ve added voiceover narration to further enhance the demonstration of these motion capture capabilities. The voiceover tutorial begins at 1:45 into the video so make sure you have your sound turned on.
To learn more about using Santos Pro, watch other videos:
https://www.santoshumaninc.com/wp-content/uploads/2019/08/Screen-Shot-2019-08-12-at-1.00.41-PM.png457803Steve Beckhttps://www.santoshumaninc.com/wp-content/uploads/2018/03/SantosHuman_Logo_Final-1-300x117.pngSteve Beck2019-08-12 13:07:142019-08-12 13:10:11Video Demo: Support for Motion Capture
In a previous blog series (post #1 of which can be found here), I had said that traditional design processes do not include human-in-the-loop (HITL) evaluations until so late in a product’s development cycle that change is no longer a realistic option. While that series focused on just four examples, each from a different industry, SantosHuman Inc. (SHI) works with clients from many industries and the anecdotal evidence for that statement is overwhelming.
Blog series posted. Next topic. Moving on.
Or at least, I thought so until a recent client engagement provided an example so perfectly suited to that series that I found myself agitated that it wasn’t included. This post alleviates that agitation and, going forward, similar posts could serve as addendums to that series. However, these addendum posts would be used sparingly and only for the most relevant examples because our experience at SHI indicates examples will continue to be plentiful.
In this first addendum entry, a global manufacturer in a highly competitive industry redesigned the thumb-operated switchgear on a new product scheduled to be in production soon. They told us their usual approach to evaluating something like this requires a working physical prototype and can take between two and three months to complete. Early on they were notified that a prototype would not be available prior to production so their plan had been to perform the study as soon as the vehicle was in production. The problem was that the company expected to be producing thousands of these vehicles every day when launched.
Let’s review that last paragraph just in case one of your eyebrows isn’t now several centimeters higher than the other. If the evaluations of the newly designed switchgear were to proceed as originally planned, tens of thousands of their products would be in dealerships and in the hands of new owners by the time the evaluations were complete. As you can imagine, the thought of having to address usability issues for tens of thousands of vehicles, coupled with the loss of market share that accompanies poor operator feedback in a highly competitive market, was causing considerable concern.
They contacted us hoping there might be some way to use just the CAD geometry to perform the switchgear usability evaluations prior to launch.
Once again, we see that traditional and pervasive approaches to product design not only value capability over usability, existing processes make it impossible to even consider usability until it’s too late to do anything about it.
The good news? As a result of development efforts SHI undertook in 2014 to respond to the needs of one of the world’s largest consumer goods manufacturers (which you can read more about here*), Santos® capabilities include the ability to predict precision and power grasps. As with all Santos® predictive models, this is a 1st principles approach to predicting grasps that not only affects – and is affected by – the entire body, it can simultaneously take into consideration other competing operator task requirements.
Our clients aren’t interested in replicating what they can already do with the virtual mannequins that have been around for decades. SantosHuman Inc.’s clients are looking for solutions to problems that would be impossible to solve without us.
https://www.santoshumaninc.com/wp-content/uploads/2019/05/PrecisionGraspingImage.png8791149Steve Beckhttps://www.santoshumaninc.com/wp-content/uploads/2018/03/SantosHuman_Logo_Final-1-300x117.pngSteve Beck2019-05-20 07:58:372019-06-24 14:38:30The Expense of Not Optimizing the HITL: Predictive Models for Precision Grasps
As you might expect, I spend a lot of my time talking to people about Santos® technologies. When they are familiar with digital human modeling at all, they’ll often say things like, “Digital human models have been around for decades. Our teams have tried them but feel they’re difficult to use and ultimately not that much of a value-add. Why should we be interested in yours?”
For those of us who have been involved from the very beginning in what is now commonly referred to as Santos®, the answer is obvious. But, simply saying, “Santos provides the ability to predict human physical behavior and performance“, isn’t meaningful before also providing a great deal of additional background information. This blog post attempts to make one of the many significant values of this unique capability a bit more obvious.
The video linked to below provides a side-by-side comparison highlighting the difference between using a truly predictive human model (on the left) versus the way in which a more traditional digital mannequin is used (on the right). While Santos® predictive models provide significant advantages for human-centric design and evaluation in any industry, this video focuses on a contrived cab space development application.
How to Watch the Video Both the left and right sides of the video were created using a single digital human character within our flagship product, Santos Pro. The right side of the video mimics the traditional way in which digital human models were designed to be used. The left side demonstrates the use of Santos predictive models.
The right side of the video only needs to be watched once through the first iteration. There’s a lot happening on the right side of the video at first so it’s not only initially more interesting, it’s almost impossible not to watch. In comparison, the activities on the left side are rather boring at first as the user is just setting up the constraints required to define an operator task. So go ahead and focus on the right side through the first iteration. The activities shown on the left will complete at about the same point in the video as the 1st iteration of the activities on the right so you’re not going to miss anything. Note, however, that the two sides only complete at about the same time because the right side has been sped up by about 5x and that’s an important point. It takes less than a minute to set up the predictive model task on the left but takes about 5 minutes for a highly experienced, expert user to manually rotate individual joints into position on the right.
After the first iteration of activities on the right is complete, that clip just repeats over and over until the end. But you’ll find you don’t have to watch the right side very long to see that manually rotating digital mannequin joints is non-intuitive, time-consuming, and tedious. In addition, it is clearly a highly subjective process where compromised, even non-human-looking, results are a frequent option. And as if that wouldn’t be frustrating enough, consider there are no economies of scale to using a digital mannequin. Every design option explored requires another round of subjective and tedious manual joint rotations. And then, after all that effort, when you’re all done, what is it you actually know? It’s no surprise that many design teams consider the use of digital human mannequins an obstacle as opposed to a solution that can be used to bring better, customer-focused designs into production sooner.
In contrast, use of a truly predictive human model (the left side of the video) allows multiple and even competing task objectives to be evaluated in a system-of-systems approach that, in this example, includes seat location, steering wheel and pedal use, and even a vision requirement. The advanced predictive nature of Santos enables your teams to identify Human-in-the-Loop requirements at the earliest stages of product development while change is still a cost-effective option.
SantosHuman Inc. When getting it wrong is not in the budget.
Take a look and let us know what you think. We’d love to hear from you.
https://www.santoshumaninc.com/wp-content/uploads/2019/04/Screen-Shot-2019-04-11-at-9.45.29-AM.png500922Steve Beckhttps://www.santoshumaninc.com/wp-content/uploads/2018/03/SantosHuman_Logo_Final-1-300x117.pngSteve Beck2019-04-15 08:36:542019-04-15 06:55:03Video Demo: What's the Difference Between a Predictive Human Model and a Traditional Digital Human?
The true value of Santos® predictive human models is the ability to provide the trade-off analysis your design teams can actually use at the earliest stages of product development. In this brief video demo, we show you how Santos® technologies can take the guesswork out of human-centric design through a contrived example focusing on seated operator sight lines. Task-focused prediction of human physical behavior and performance makes this possible in minutes instead of hours, weeks, or even months of trial and error and physical prototypes.
Let us know what you think. We’d love to hear from you.
If you want to learn more about how to use Santos Pro® or Santos Lite®, our new video series can help. We’ve introduced the first of four Master Class Series videos on our YouTube channel. The first video covers how to predict and simulate task-focused physical human behavior and performance. The second video tutorial shows how Santos Pro avoids collisions and vision obstructions.
Please check back for more videos to be added to this series of tutorials, and contact us if you would like to learn more about Santos Pro® and its capabilities.
https://www.santoshumaninc.com/wp-content/uploads/2019/01/Screen-Shot-2019-01-15-at-9.44.06-PM.png444650wedadminhttps://www.santoshumaninc.com/wp-content/uploads/2018/03/SantosHuman_Logo_Final-1-300x117.pngwedadmin2019-02-11 12:00:372019-02-18 09:16:04Introducing The Santos® Master Class Series on YouTube