The Washington PostDemocracy Dies in Darkness

This robot quarterback could be the future of football practice

The machine, called the Seeker, saw its NFL debut this week with a test run by the Green Bay Packers

Green Bay Packers safety Micah Abernathy catches a pass during NFL football training camp with the New Orleans Saints on Wednesday. (Samantha Madar/AP)
5 min

When the Green Bay Packers walked onto the practice field this week, they were greeted by an unusual new teammate: a robot.

In videos on Twitter, a 6-foot-tall white robotic machine simulates a punter, kicking balls at a rapid pace to players downfield. The robot, which holds six balls in a revolving cartridge, could also imitate a quarterback’s style including the speed, arc and timing of a throw.

The Seeker is a robotic quarterback, kicker and punter rolled into one. It’s a modern day version of a piece of football equipment, called a JUGS machine, that’s been used to simulate throws and kicks to football players for decades. The Seeker, company officials say however, is a more accurate thrower and runs software to let players practice more advanced gameplay scenarios.

The Seeker, created by Dallas-based Monarc Sports, is a robotic quarterback, kicker and punter rolled into one. (Video: Monarc, via Youtube)

The robot, created by Dallas-based Monarc Sport, is starting to gain adoption. Top college football programs, such as Louisiana State University, the University of Oklahoma and the University of Iowa, all count the Seeker as part of their training strategy. The Green Bay Packers are the first team in the National Football League to try the technology.

Google is training its robots to be more like humans

The Seeker’s software allows players to customize how they practice with it. Athletes can catch balls from close to the machine to improve hand-eye coordination. They can also program the robot to throw a ball to a spot on the field, or simulate more-lifelike conditions by over or underthrowing a ball. Players wear a pager-like tag which allows the robot to track their location on the field, and throw a ball accurately within inches.

“It gives so much opportunity for our guys to get reps without the need of having a quarterback there,” said Ben Hansen, the director of football administration at Iowa, where the technology was first tested. “That’s a huge plus.”

Since the 1970s, football teams have relied on the JUGS machine to avoid wearing out quarterbacks and kickers. It fired off footballs through two high-speed rotating discs and allowed players to run routes or practice catching by themselves, operating with basic machinery and not utilizing software.

Over the decades, the machine — named after its creator, JUGS Sports — became commonplace on the football field. But it’s been criticized by football staff members for poor performance.

Matt LaFleur, the head coach of the Green Bay Packers, spent a few minutes in early August criticizing the JUGS machine for not simulating punts well. “It was awful,” he said in a news conference. “You couldn’t get the ball to turn over. It was damn near impossible to catch.”

Lifeguard drones can save lives

J.R. Reichenbach, a national account manager at JUGS Sports, said the company contacted the Packers after seeing the clip to ask if they could help alleviate the issue. “We were there for them,” he said. “They didn’t need anything, everything’s fine.”

Igor Karlicic and Bhargav Maganti, co-founders of Monarc Sport, started working on the Seeker in 2015 as former engineering students at Northwestern University looking for a way that wide receivers could train on their own. They created a prototype and worked with Iowa to refine the concept.

The Seeker robot has two rotating discs, similar to a JUGS machine, that rotate quickly and help launch a ball. The robot can carry six balls at a time in cartridges, similar to a revolving gun chamber. Each robot costs roughly $40,000 to $50,000 per year for the hardware, software and servicing, Karlicic said.

“Small advantages matter a lot,” Karlicic said of the training options the robot offers. “All of this makes a huge impact on game day.”

Hansen, of Iowa, said in an interview that his team started using the Seeker in 2018. One of the most helpful parts of the technology, he said, is being able to program it to throw passes that simulate game day conditions. Unlike the JUGS machine, he said, which doesn’t have software to pass in random patterns, the Seeker can purposefully throw passes that aren’t perfect.

“Every single pass isn’t always going to just hit you in the chest,” he said. “So to be able to practice and simulate different types of passes that are coming at you only help your ability to be more efficient and productive with regards to catching.”

A case study published in April by Microsoft, which provides the software ecosystem for the robot, noted that West Virginia University’s dropped pass rate fell to 4 percent in 2021, after introducing the robot into training, down from 8.5 percent the past season. The university’s senior athletic director said the robot deserved a “share of the credit” for that outcome.

These robots were trained by AI. They became racists and sexist.

After LaFleur complained about the Packers’ JUGS machine, Karlicic said his company fast-tracked a plan to give the team a trial of the robot which was at practice this week. The team is not an official customer yet, Karlicic said, but has been conversing with Monarc Sport for months.

The Green Bay Packers declined to comment.

Daron K. Roberts, a former NFL assistant coach and director of the University of Texas’s Center for Sports Leadership and Innovation, said in an interview that he is not surprised football teams are interested in the Seeker. In recent years, the NFL has been looking to wearables, drones and other forms of technology to automate coaching and team operations.

“Technology has infiltrated the NFL,” he said.

Roberts said most teams in the NFL will likely look at the Packers trial run to gauge if it has any tangible benefits.

“The NFL is a very copycat league,” he said. “If another team has an edge, other teams are going to follow.”


A previous version of this article misstated West Virginia University’s dropped pass rate. It was 8.5 percent before the robot was introduced, not 53 percent. The article has been corrected.