#AU2016 Opening Keynote

Autodesk CEO, Carl Bass, and CTO, Jeff Kowalski, kicked off Autodesk University 2016. They talked about the trends and technologies they think will have the biggest impact on the work our customers do and presented exciting customers from around the world who are doing some very innovative things.

Here are my takeaways from Jeff's portion of the opening keynote.

• Artificial Intelligence and Machine Learning

  The most important thing happening in software today is the acceleration of Artificial Intelligence and Machine Learning. We're bringing machine learning to the 3D world. We're feeding our algorithms huge amounts of 3D model data so they can grasp the essence of the designs we work on every day. With that new understanding, the software can take a 3D object, like a generic chair for instance, and apply a specific style to it to design a new chair.

• Generative Design

  Generative Design is a new way of collaborating with a computer where the designer doesn't tell it what to do. Instead, the designer tells it what he needs. Our plan is to release our Generative Design tools commercially next year.

• Augmented and Virtual Reality

  Creating a great experience for someone is really hard unless you, as the designer, can also experience that thing while you're creating it. Now, we can finally do this by immersing ourselves in the digital realm through Virtual Reality, where we can experience our ideas first hand before they're even real. Autodesk LIVE can take a Revit model into a VR environment with a few clicks, and open the doors to that building for all to experience.

• Robots and Sensors

  When you team up robots with Machine Learning and Generative Design, that combination enables infinite expressibility. By giving robots senses and an awareness of their human collaborators, humans and robots can work together, side by side, to do things that neither could do on their own. Thinking about technologies like robots as a threat is 100% wrong. Technologies aren't a threat — they're more like superpowers.

Here are my takeaways from Carl's portion of the opening keynote.

• Automotive Industry as an Example

  There are three things transforming the auto industry: autonomous cars, car sharing, and electric power trains.

  For a hundred years, car companies have been obsessed with creating a great experience for drivers. But when cars drive themselves, there's no such thing as a driver experience. Instead, car companies need to build the ultimate passenger experience. They need to get good at building complex sensing and control systems. Their new cars will need software that responds to the world in milliseconds. They'll have to get smarter the more they drive, so they don't make the same mistakes twice. These companies can't just be building cars — they need to start building drivers.

  Not only will we stop driving our cars in the future, we won't own them either. Most people under 30, don't own a car, and they don't want to. Now car companies have to figure out how to thrive in a new world where people don't pay to buy cars but pay to access them instead.

  Traditional auto companies were completely dismissive of electric cars. They made fun of them. They said electric cars were a crazy California idea that wasn't going to go anywhere. They spent years in denial, and now they are struggling to catch up.

• Machine Learning as a Software Approach

  In the past, we used to write some code, and the program would do exactly what it was told. The only way to make that program better, other than re-writing it, was to put it on a faster computer. Today, this kind of deterministic software programming is being replaced by machine learning. For instance, now we can use algorithms to understand the work designers are doing, and monitor how designers use our software. Then the software can customize itself to become exactly the tool a designer needs. This is already real. It's available now in Design Graph on A360. We can recognize the design being working on based on its shape and show the designer all the other things that look like it, or relate to it. If a designer has a bolt in his design, it will show him nuts and washers. It can also make good guesses about what the designer going to need next, and point him to designs that already exist, so he doesn't re-invent the wheel.

  One of the biggest headaches for machinists is figuring out the cutting parameters they need to use for programming their machines. Speeds and feeds. Master machinists spend their entire careers honing their skills in this dark art. So this summer, some folks at Autodesk figured out how to use Machine Learning to automate speeds and feeds. So now, engineers don't need to worry about that anymore. The algorithm just does it for them.

• Teams

  We're not building tools for individuals any more. Instead, we're building tools for teams. Like the car companies, we have to look past our own expertise (single-user desktop applications) to a new generation of tools that make teams more productive. When people communicate changes, anticipate needs, and respond together to a changing environment, the outcome is artful and coherent. We're already building tools that facilitate coordination, collaboration, and communication of work like BIM 360, Fusion 360, and Shotgun. These tools were built around team effectiveness.

  Our cloud services act on behalf of our customers and their teams, without being told. Updates get propagated. Renderings get generated. Notifications are sent. Simulations are kicked off. Results are published. Versioning is preserved. Like a trusted team member. These services are autonomous and reliable. And they're not tying up the user's desktop computer, so the throughput of the whole team is improved.

• Shared Data

  We are building tools around shared data, so we no longer have our own versions of reality any more. We have shared Revit models and shared Fusion 360 models. When there's a single source of truth, and you ask the structural engineers: "Is this change OK?" he isn't so busy finding the design files and pulling them across the wire that it takes them a week to respond.

It was a great way to start off #AU2016.

Kickoff activities are alive in the lab.