Innovation the “DARPA Hard” Way

By July 30, 2019March 30th, 2021No Comments

At IAM Robotics, we are interested in addressing complex and technically difficult problems, the solutions for which will disrupt the supply chain industry and create better, more productive e-commerce fulfillment processes. Specifically, we are focused on autonomous mobile picking, which combines navigation, perception, and manipulation—three of the hardest things to do together in robotics. We have successfully done this, creating Autonomous Mobile Manipulation Robots that have the potential to completely transform the logistics industry. 

Our desire is to create innovation, which goes beyond invention by turning it into a practice that requires a fundamental change in operations. We call this type of innovation, “DARPA Hard.” The culture of IAM Robotics is born from the desire to tinker, to experiment, and to push beyond what we thought was possible. This is transformative innovation, and this is the culture of IAM Robotics. 

IAM Robotics was founded by Tom Galluzzo and Vladimir Altman, who met while working together at the National Robotics Engineering Center (NREC) at Carnegie Mellon University. Prior to his time at NREC, Tom participated in the DARPA Grand Challenge Project in 2004 and 2005, while in graduate school. After joining NREC, he worked on the DARPA Autonomous Robotic Manipulation-Software (ARM-S) Project.

He is now leading the development of the Swift robots to do new warehouse tasks and to deploy the robots on a larger scale for bigger customers.

The Defense Advanced Research Projects Agency

Over its almost six decades of existence, the Defense Advanced Research Projects Agency (DARPA) has supported and guided work that has “changed the world” by creating transformative innovation as opposed to incremental improvements in existing technologies. The truth is, DARPA was born out of crisis, a reaction to the Soviet Union’s launch of the Sputnik satellite in 1957 that took the United States by surprise. Founded in 1958 under the Eisenhower administration, DARPA is  the central research and development organization of the Department of Defense (DOD) and has a very specific mission: innovation in the defense of our country. 

DARPA has played a role in the development in many of the most transformative technologies that exist, including the Internet (initially known as Arpanet), stealth aircraft, miniaturized GPS technologies, unmanned aerial vehicles, flat-screen displays, and the brain-computer interface work that is making it possible for subjects to use their thoughts to move artificial limbs. 

DARPA seeks challenges that are aggressive enough to create energy and vision and require a stretch to get there, without going beyond the possible. It’s about asking the “Why? What If? and How?” questions that are ambitious and actionable, and then encouraging innovation that answers these questions but allows for failure as a means to move technology forward. 

Photographed above is Tom (with more hair on his head and none on his face) when he participated with the University of Florida team that responded to DARPA’s Grand Challenge for self-driving cars. This path of discovery initiated by DARPA has left everyone thinking about cars differently.

What Is “DARPA Hard”?

DARPA Hard ideas are high-risk and high-reward – often considered crazy – that when accomplished have the ability to challenge an entire industry or catalyze the formation of a new one. They are big visions, well beyond current capability, and oftentimes sit at the intersection of multiple fields. “DARPA Hard” ideas are far-reaching, technically challenging, multi-disciplinary, and actionable.   

The four dimensions addressed in the DARPA Hard concept.

Source: Lockwood Resource

While each dimension is important in its own right, it is the combination that matters most and leads to truly culture changing and visionary innovation.  The DARPA Hard Test assesses an idea across the four dimensions of DARPA Hard, and scores each dimension on a scale of 1-7, with 7 being the highest score.

Pushing Innovation at IAM Robotics

We are driven by our desire to make flexible automation a reality, and utilize the “DARPA Hard” framework in order to continue pushing innovation that will ultimate disrupt the supply chain industry:


In the past, the repetitive and mundane tasks necessary for order fulfillment was done by humans. As e-commerce continues to grow, along with increasing customer expectations of nearly immediate delivery, there has been a necessary paradigm shift that recognizes and embraces that at least some of these tasks can be completed with autonomous robots. And while previous attempts at automation technology required non-human-friendly infrastructures (gantry structures, cages, etc.), robots are now able to work within a structure that is flexible, and familiar to and comfortable for warehouse workers. 

Beyond what other autonomous robotic solutions offer, the Swift Product Suite offers a unique combination of navigation, perception, and manipulation to provide for a complete order fulfillment solution. The Swift robot travels autonomously into warehouse storage areas, finds the inventory locations, individually identifies objects, and picks the objects. This allows the fulfillment workers to increase their productivity by not having to walk around the facility to get things. Working with a team of robots makes warehouse workers more efficient, and allows the warehouse to sell more items. The end result is that the fulfillment company can make more money and provide a much greater quality of life for its workers. 

Technically Challenging

Robots have been classically blind, but we are using computer vision and other AI technologies to make them interact within their environment. Anyone can put an off-the-shelf industrial robot arm on a wheeled base, but it’s very hard to develop the algorithms necessary to coordinate the two capabilities. Thus, the key to making completely autonomous mobile manipulation robots work is vision. Specifically, vision sensor and perception algorithms  that allow the robot to adjust for things that are positioned out of place, including the robot itself. The sophisticated integration of perception sensors, autonomous navigation, manipulation, and artificial intelligence in Swift warehouse robots enables maximum picking productivity across single picks, trays, and totes. 


Robotics is one of the most multi-disciplinary fields in engineering today. Building robots require mechanical engineering; computer science, including AI, computer vision, and software engineering; electrical engineering; and particularly important for us at IAM Robotics, industrial engineering.  And beyond building the robots, deploying the robotic workforce, and working with our customers throughout the entire process is done in a multi-disciplinary and holistic manner, as well. 

From early on in the sales process, our team works to gain a full understanding of what our clients want to achieve and accomplish in their warehouse automation. These discussions include specifics about their demand profiles, current operations, product attributes, upstream and downstream systems, and more. The IAM Robotics team then pulls together models and simulations to help clients visualize how the solution will impact their business operations. Once the Swift Product Suite is deployed into a client’s operations, we have regular meetings to discuss client innovation improvements, preventive maintenance, and new features and capabilities deployment schedules.


While we are continuously striving to push innovation, we fully understand what our robots can and cannot do today. Our Swift robot is particularly suited to pick consumer-packaged goods for retail environments including center-aisle grocery and drugstore retailers. We continuously work with our clients to see how we can increase the products that our robots can pick. Additionally, we assess how and where the products are placed within the client’s warehouse and how modifications to the set-up can provide more efficiencies. 

While the number of products that a robot is able to pick certainly attributes to the overall efficiency of the warehouse, the true success rate for a robotic workforce lies in its ability to provide a continual stream of uninterrupted work over time. 

The Autonomous Revolution

Ultimately, the greatest challenge facing many companies is how to compete better in e-commerce and improve the customer experience without investing high levels of capital in transforming their distribution and fulfillment infrastructure. IAM Robotics has  created proprietary Autonomous Mobile Manipulation Robots that empower companies to transform their fulfillment for the on-demand economy.

Just as DARPA projects seek to invest in platforms that will kickstart new industries and continue to build on initial breakthroughs, IAM Robotics is looking to create an autonomous robot revolution that will change how the logistics industry operates. Our groundbreaking autonomous robot solutions improve e-commerce efficiency, giving retailers a competitive advantage, and empowering them to solve today’s operational challenges. IAM Robotics transforms e-commerce fulfillment centers for the on-demand economy with innovative Autonomous Mobile Manipulation Robots capable of fully automating order picking and retrieval in warehouses designed for human labor.