On the Fringe: Creating autonomous vehicles in our homes, offices, factories and hospitals

In our ongoing ‘On the Fringe’ series, some of TI’s brightest minds discuss today’s biggest technological trends and solving the challenges of tomorrow.

When I tell people that I am the director of autonomous vehicles research and development (R&D) at TI Kilby Labs, the response I usually get is, ‘Like cars that drive themselves? Cool!’ And, yes, we are doing some amazing things here at TI to enable autonomous vehicles.

While we do focus a lot on driverless cars, trucks and SUVs, autonomous vehicles expand well beyond our roadways. We’re taking the same technology being developed for cars and putting it into all types of vehicles and robots. Our future will be much more autonomous than you can ever imagine.

For example, we already know the unmanned aerial vehicles (UAVs) market is huge, with UAVs being used in all sorts of applications, but making these vehicles not just unmanned, but autonomous, is the next step. Autonomous UAVs can be used in military applications, to distribute goods from online retailers, for video recording at sporting events, TV shows and movies or even as surveillance and security for businesses and homes. Imagine two quadcopters (a form of a UAV) where one sits on a battery charging station while the other flies around your property, monitoring for and focusing on moving objects. The quadcopter can identify a moving object (like a criminal) and quickly warn of an intruder. Once its battery runs low, the first quadcopter flies back to the battery charging station and the other quadcopter takes off without any human interaction.

On the medical front, researchers are experimenting with robots treating children with Autism or other developmental disorders. The robots must be able to autonomously sense the behavior of a child and respond or react accordingly. Use of robots exist today in hospitals, and I expect this trend to continue to grow, although I believe we are still a long way away from robots performing surgeries without human interaction as there are many dexterity issues that still must be resolved.

 From construction sites to battlegrounds to farms, vehicles on four wheels are the backbone of these industries. But creating autonomous vehicles for building a skyscraper or to plow a field of cotton has many more complications to overcome than vehicles on the road. Roadways are very structured, from the material used to pave roads to lane markings on the road. Off the road, the environment is very unstructured, with no lanes to guide vehicles and issues such as whether the ground beneath the vehicle is hard or soft. Essentially, there are many more types of obstacles, so it will take us longer to create systems that can identify all of these challenges. But I believe these types of vehicles will become a reality in the future.

When it comes to autonomous robots in our homes, most do nothing more than clean our floors today – but I believe they can work smarter and better, with increased capabilities, than what we have available now. And why couldn’t an autonomous vehicle/robot fold or iron our clothes? Why can’t a robot load the dishwasher? There are near limitless applications inside the household for chores to be completed autonomously.

Autonomous robots exist to some degree in warehouses, but there is a great opportunity to create vehicles that have the capabilities to find and pull objects from a shelf in a massive warehouse and then package the objects for shipment.

The final area where I see autonomous vehicles is in an office setting. Telepresence vehicles exist today where with a joystick or controller a person can remotely move about an office with a fixed camera and monitor. But we hope these telepresence vehicles can one day autonomously find office rooms simply by saying into a microphone, for example, ‘Go find Daniel.’ The vehicle will know where Daniel’s office is, go to Daniel’s office door, knock, and open the door when Daniel gives the proper response (like, ‘Come in.’).

While the applications are nearly endless for autonomous vehicles, how do we go about making them a reality? Autonomous vehicles are a two-step process. First, these vehicles must be able to understand the world around them. Using cameras, radar, lidar and ultrasound, vehicles can get a very good 3D picture of objects near and far. The second step involves translating that data into something a computer can understand. This processing takes place in a microprocessor that can then tell the vehicle about its environment and how to respond, whether grabbing a particular item off a shelf, record the trespasser on your property or spend a little extra time getting the wrinkle out of a shirt while on the ironing board. Once a vehicle understands the environment around it, then it can focus on the task at hand.

This technology is available today, but at a great cost. My job is to make sure this technology continues to develop and improve so that one day it can become cost effective, and vehicles and robots all around us, in all shapes and sizes, will help run our world autonomously.

If you want to learn more about autonomous vehicles like cars and all the ones described above, read this white paper and watch the video below with my Kilby Labs colleague Kristen Parrish discussing all the ways we’re enabling the autonomous vehicles of the future.