CtrlK
 | |
| Developer | Kinisi Robotics |
| Type | Humanoid robot (wheeled) |
| Also known as | Kinisi 01, KR1 |
| Unveiled | July 2025 |
| Country of origin | United States / United Kingdom |
| Height | 162 cm (5 ft 4 in) |
| Weight | 100 kg (220 lb) |
| Degrees of Freedom | 21 |
| Mobility | Omnidirectional wheeled base |
| Top Speed | 2.4 m/s (8.6 km/h) |
| Payload | 25 kg dynamic / 40 kg static |
| Battery | 48V Li-ion, hot-swappable; 6 to 8 hours runtime |
| Computing | NVIDIA Jetson |
| Sensors | Stereo depth cameras, 180-degree LiDAR, SLAM |
| Price | ~$75,000 USD (or ~$4,000/month rental) |
| Status | In production; pilot deployments |
| Website | kinisi.com |
The Kinisi KR1, also marketed as the Kinisi 01, is a wheeled humanoid robot developed by Kinisi Robotics for warehouse automation, logistics, and light manufacturing. Unlike most humanoid robots that rely on bipedal locomotion, the KR1 uses an omnidirectional wheeled base to prioritize speed, stability, and practicality over human-like walking. Standing 162 cm tall and weighing 100 kg, the robot features dual arms with modular grippers, an IP65-rated aluminum chassis, and on-board artificial intelligence powered by an NVIDIA Jetson module. Kinisi Robotics positions the KR1 as an affordable and deployment-ready alternative to more expensive bipedal humanoids, with a list price of approximately $75,000 and a rental option at roughly $4,000 per month.
The KR1 was first publicly unveiled in July 2025 after roughly 18 months of development. The robot won the Automate Start-up Challenge in May 2025, and by November 2025 it had entered its first commercial deployment at a glass recycling facility in the United Kingdom. Kinisi Robotics has also secured a pilot agreement with a major global automotive manufacturer for intra-logistics tasks.
Kinisi Robotics is a robotics startup with offices in New York City (140 Broadway, New York, NY 10005) and Bristol, United Kingdom (Beacon Tower, Colston Street, Bristol BS1 5AQ). The company was formally established in January 2024 by Brennand "Bren" Pierce, a roboticist with over 15 years of experience in robotic hardware design and software development. Pierce previously co-founded Robotise GmbH (a Munich-based startup that created autonomous mobile delivery robots for hotels) and Bear Robotics (a Silicon Valley company that developed the Penny restaurant delivery robot and raised $32 million in Series A funding led by SoftBank). Across his career, Pierce has been involved in deploying over 10,000 robots worldwide and raising more than $180 million from investors.
Pierce holds a PhD in humanoid robotics from the Technical University of Munich (TUM), where he conducted research at the Institute for Cognitive Systems from 2009 to 2015. During his academic career, he built a full-size compliant humanoid robot called "Herbert" and developed a new generation of FPGA-based robotic controllers. He also worked on a project with Samsung at Carnegie Mellon University. His professional philosophy centers on building robots that are economically viable for real-world commercial applications rather than research demonstrations.
The name "Kinisi" derives from the Greek word for movement, reflecting the company's focus on creating technology that makes physical work more efficient.
As of early 2026, the Kinisi Robotics leadership team includes:
| Name | Role | Background |
|---|---|---|
| Brennand Pierce | Founder and CEO | PhD in humanoid robotics (TUM); co-founded Robotise and Bear Robotics |
| Aaron Colfer | Head of Product | Former Senior Manager of Industrial Design at Dyson (2021 to 2024); Senior Design Lead at Matter |
| Mitch Leibowitz | Head of Business Development | Business development and partnerships |
| Mark Finean | Head of Robotics | Robotics engineering lead |
Kinisi Robotics has raised approximately $2 million in seed funding as of early 2026. The company's development has followed a rapid timeline:
| Date | Milestone |
|---|---|
| January 2024 | Company founded; initial customer conversations begin |
| May 2024 | KR1 prototype completed with agentic AI capabilities |
| September 2024 | First employees hired; Bristol office established |
| October 2024 | First customer visit |
| April 2025 | V2 hardware completed |
| May 2025 | First on-site customer demonstration in Europe; won Automate Start-up Challenge |
| July 2025 | Public unveiling of the KR1 platform |
| September 2025 | V3 hardware launched |
| November 2025 | First commercial deployment (glass recycling in UK) |
The KR1's most distinctive design choice is its use of a wheeled base instead of bipedal legs. Founder Bren Pierce has been vocal about this decision, arguing that legs are unnecessarily complex for the vast majority of industrial applications. In an interview with Tech.eu, Pierce explained: "Customers don't necessarily want that. In our conversations with warehouse and factory operators, they see humanoids with legs and hands as cool toys, not industrial tools." He has also questioned the engineering trade-offs involved in bipedal locomotion, asking: "Why spend on 14 motors when two will do?"
This philosophy extends to the company's broader strategy. Kinisi focuses on what it calls "software layering," building robust AI software on top of proven, reliable hardware rather than trying to simultaneously reinvent both hardware and artificial intelligence. Pierce has drawn comparisons to past failures in the consumer robotics industry, citing companies like Anki, Jibo, and Kuri as cautionary examples of products that over-engineered hardware without achieving practical commercial viability.
The KR1 targets what Kinisi describes as the 70% of warehouses that are small, irregular spaces unable to justify massive fixed automation investments. Pierce has noted that these facilities "can't justify huge automation budgets. Our robots are attractive here because they don't require cages." The robot is designed to integrate into existing workflows without requiring new infrastructure, conveyor systems, or facility redesigns.
The KR1 stands 162 cm (approximately 5 feet 4 inches) tall and weighs 100 kg (220 lb). The frame is constructed from aluminum and carries an IP65 ingress protection rating, meaning it is fully protected against dust and low-pressure water jets from any direction. This level of protection allows the robot to operate in industrial environments where dust, debris, and occasional liquid splashes are common. The chassis is precision-engineered with gaps no larger than 0.5 mm, and joints are protected with polyurethane bands to prevent contaminant ingress.
| Category | Specification | Value |
|---|---|---|
| Physical | Height | 162 cm (5 ft 4 in) |
| Physical | Weight | 100 kg (220 lb) |
| Physical | Frame material | Aluminum |
| Physical | IP rating | IP65 |
| Physical | Chassis gap tolerance | 0.5 mm or less |
| Mobility | Base type | Omnidirectional wheeled |
| Mobility | Turning radius | Zero-turn capable |
| Mobility | Top speed | 2.4 m/s (8.6 km/h / 5.4 mph) |
| Mobility | Degrees of freedom (total) | 21 |
| Mobility | DOF per hand | 2 |
| Manipulation | Dynamic payload | 25 kg (55 lb) |
| Manipulation | Static payload | 40 kg (88 lb) |
| Manipulation | Gripper type | Dual-end gripper with quick-swap interface |
| Manipulation | End-effector options | Mechanical fingers, vacuum tools |
| Actuators | Motor type | Brushless DC (BLDC) |
| Actuators | Gear technology | Strain wave (harmonic drive) |
| Actuators | Arm modules | Compact wave gear with integrated encoders |
| Power | Battery voltage | 48V |
| Power | Battery type | Lithium-ion (20 Ah modular pack) |
| Power | Runtime | 6 to 8 hours (typical industrial duty cycle) |
| Power | Charging | 80% in 90 minutes (CC/CV smart charger) |
| Power | Battery swap | Hot-swappable with auto-docking |
| Power | Battery management | Integrated BMS with thermal and current protection |
| Perception | Cameras | Stereo depth cameras |
| Perception | LiDAR | 180-degree array with SLAM integration |
| Perception | Depth accuracy | Plus or minus 2 mm at 2 meters |
| Perception | Perception-to-action latency | Sub-100 ms |
| Computing | Processor | NVIDIA Jetson (CPU/GPU) |
| Computing | AI inference | Real-time transformer models |
| Computing | Operating system | Linux |
| Connectivity | Cloud telemetry | Encrypted, via Kinisi Mission Control |
| Connectivity | Updates | Over-the-air (OTA) firmware |
| Connectivity | API | Available |
| Connectivity | ROS compatible | Yes |
| Safety | Controllers | Dual redundant with watchdogs |
| Safety | Emergency stop | Hardware and software linked |
| Safety | Diagnostics | Continuous thermal and current monitoring |
| Safety | Fail-safe | Automatic deceleration and posture lock on critical alerts |
The omnidirectional wheeled base provides zero-turn capability, allowing the KR1 to rotate in place and navigate tight aisles and confined warehouse spaces. The base includes an active damping suspension system that compensates for terrain variation and load shifting, helping maintain stability when the robot is carrying heavy objects. At a top speed of 2.4 m/s (approximately 8.6 km/h), the KR1 moves significantly faster than most bipedal humanoid robots, which typically walk at 1.0 to 1.5 m/s. The wheeled platform also eliminates the risk of falls that bipedal systems face on smooth warehouse floors.
The KR1 features dual human-scale arms with 21 total degrees of freedom distributed across the body, including 2 DOF per hand. The arms use compact wave gear modules with integrated encoders, driven by brushless DC motors through strain wave (harmonic drive) gearing for precise, low-backlash motion.
The gripper system uses a dual-end design with a quick-swap interface that allows operators to change end-effectors depending on the task. Available options include mechanical finger grippers for grasping items like bottles and boxes, and vacuum end-effectors for handling flat or smooth-surfaced objects such as cartons and poly-bags. The robot can handle dynamic payloads of up to 25 kg and static loads of up to 40 kg.
The KR1's sensor suite combines stereo depth cameras with a 180-degree LiDAR array. The stereo cameras provide depth perception with an accuracy of plus or minus 2 mm at a range of 2 meters, which is sufficient for precise picking and placement operations. The LiDAR array feeds into a SLAM (Simultaneous Localization and Mapping) system that builds and continuously updates a map of the robot's environment, enabling autonomous navigation without pre-programmed paths.
All sensor data is fused in real time to create a 3D representation of the workspace. The perception-to-action latency is under 100 milliseconds, meaning the robot can detect an obstacle or identify an object and begin responding in less than a tenth of a second.
The KR1 runs on a 48V lithium-ion battery with a 20 Ah modular pack. Under typical industrial duty cycles, the battery provides 6 to 8 hours of continuous operation. Kinisi designed the battery to be hot-swappable, meaning operators can replace a depleted battery pack with a charged one without shutting down the robot. The system also supports auto-docking, where the robot autonomously returns to its charging station when battery levels drop below a threshold.
Rapid charging reaches 80% capacity in approximately 90 minutes using a CC/CV (constant current / constant voltage) smart charger. An integrated battery management system (BMS) provides thermal and current protection, monitoring cell temperatures and preventing overcharge or deep discharge conditions.
All of the KR1's artificial intelligence processing runs locally on the NVIDIA Jetson module. This edge computing approach means the robot does not depend on a cloud connection for real-time decision-making, which improves response times and ensures reliability in environments with poor or unstable internet connectivity. It also addresses data privacy concerns, as sensitive warehouse inventory data and operational patterns stay on the robot rather than being transmitted to external servers.
The AI system runs real-time transformer inference models for both navigation and manipulation tasks. The robot uses a reinforcement learning controller that processes fused sensor data (LiDAR and stereo vision) to plan paths, avoid obstacles, and adjust arm trajectories dynamically.
One of the KR1's key features is its use of demonstration-based imitation learning. Rather than requiring traditional programming or coded task sequences, operators can teach the robot new tasks by physically demonstrating them. The robot observes the demonstration and learns to replicate the task independently. Kinisi calls this approach part of its broader goal of making robots accessible to warehouse workers who may not have programming expertise.
The system also supports what Kinisi calls "Shadow Play," a teleoperation mode where a human operator can remotely control the robot's movements. Data collected during teleoperation and demonstrations feeds into the robot's machine learning pipeline, allowing it to progressively improve its autonomous performance on trained tasks.
A central element of Kinisi's software strategy is its "data flywheel" architecture. Every deployed KR1 unit contributes operational data to a shared knowledge base. When one robot learns to handle a new object type, navigate a new obstacle pattern, or optimize a picking sequence, that knowledge can be distributed to the entire fleet. This means that as Kinisi deploys more robots, each individual unit benefits from the collective experience of all deployed systems.
Fleet management is handled through Kinisi Mission Control, a cloud-based telemetry and management platform. Mission Control provides real-time monitoring of robot status, task progress, and fleet-wide analytics. It supports remote task updates, over-the-air firmware updates, and encrypted data transmission. Operators can switch individual robots between fully autonomous, semi-autonomous, and teleoperated modes through the platform.
The KR1 supports three operating modes:
| Mode | Description |
|---|---|
| Autonomous | The robot operates independently using its onboard AI, navigating, picking, and placing without human intervention |
| Semi-autonomous | The robot handles routine operations autonomously but requests human guidance for edge cases or unfamiliar situations |
| Teleoperated (Shadow Play) | A human operator remotely controls the robot's movements in real time, useful for teaching new tasks or handling unusual scenarios |
Kinisi Robotics has identified several primary application domains for the KR1:
| Application | Description |
|---|---|
| Pick and place | Retrieving items from shelves or bins and placing them into totes, boxes, or onto conveyors |
| Tote and material handling | Moving loaded totes, boxes, and pallets between stations within a facility |
| Palletizing and depalletizing | Stacking and unstacking items on pallets for shipping or storage |
| Inspection and quality control | Using stereo cameras and depth sensing to visually inspect items for defects |
| Replenishment | Restocking shelves and inventory locations |
| Assembly support | Handling components in light manufacturing and assembly operations |
The KR1's first commercial deployment took place in November 2025 at a glass recycling facility in the United Kingdom. This deployment marked the first time the Kinisi 01 platform performed live production work at a customer site. The robot was tasked with autonomous glass bottle sorting for reuse, a process that requires delicate handling to prevent breakage while maintaining throughput.
During the deployment, the KR1 used its stereo depth cameras and onboard AI to autonomously determine grasp points and classify bottles by shape, size, and material. The robot had to balance speed against careful handling, as broken glass creates both product loss and safety hazards.
Aaron Colfer, Kinisi's Head of Product, acknowledged that the system's throughput was initially below human-operator levels: "The process is still slower than a human operator at this stage. Our plan now is to gradually increase speed, with the ambition of surpassing human throughput over the next couple of months." This transparency about early performance limitations reflected Kinisi's approach of deploying real systems in production environments and iterating, rather than waiting for laboratory-perfect performance.
Kinisi has also secured a pilot agreement with an unnamed major global automotive manufacturer. The initial focus of this pilot is on intra-logistics tasks (moving materials and parts within the factory) rather than complex assembly operations. This approach aligns with the industry pattern seen in other humanoid robot deployments, where companies like BMW and Mercedes-Benz have started with material handling before exploring more complex applications.
In May 2025, Kinisi Robotics won the Automate Start-up Challenge, a competition held as part of the Automate trade show organized by the Association for Advancing Automation (A3). The challenge is a pitch competition where early-stage robotics and automation companies present their technology and business models to a panel of judges.
Kinisi's win was notable because the company was competing against other innovative startups, including 3Laws Robotics (a CalTech spinout developing a dynamic universal safety layer for robots) and Sonair (which develops ultrasonic sensor technology). The Automate Start-up Challenge provided Kinisi with industry visibility and credibility at a time when the company was preparing for broader pilot deployments. The event took place shortly before ICRA 2025 (the IEEE International Conference on Robotics and Automation) in Atlanta, Georgia.
The KR1 has gone through multiple hardware revisions in its relatively short development history:
| Version | Date | Notes |
|---|---|---|
| V1 (Prototype) | May 2024 | Initial KR1 prototype with agentic AI capabilities |
| V2 | April 2025 | Second-generation hardware; used for first customer demonstrations in Europe |
| V3 | September 2025 | Third-generation hardware; deployed in glass recycling pilot |
This rapid iteration cycle (three hardware versions in roughly 16 months) reflects both Kinisi's agile development approach and the engineering focus on getting functional hardware into real-world environments as quickly as possible. Each iteration has incorporated feedback from customer visits and pilot deployments.
The KR1 occupies a distinctive niche in the humanoid robotics market. While companies like Figure AI, Tesla, Boston Dynamics, and Apptronik are developing bipedal humanoid robots with general-purpose ambitions, Kinisi has deliberately narrowed its focus to wheeled humanoids for industrial logistics.
| Company | Robot | Locomotion | Key Focus | Price Range |
|---|---|---|---|---|
| Kinisi Robotics | KR1 (Kinisi 01) | Wheeled (omnidirectional) | Warehouse logistics, pick and place | ~$75,000 |
| Figure AI | Figure 02 / Figure 03 | Bipedal | General-purpose manufacturing | ~$100,000 (estimated) |
| Apptronik | Apollo | Bipedal | Factory and warehouse automation | Undisclosed |
| Agility Robotics | Digit | Bipedal | Logistics and material handling | Undisclosed |
| Tesla | Optimus | Bipedal | Factory automation, consumer | $20,000 to $30,000 (target) |
| Unitree Robotics | G1 / H1 | Bipedal | General-purpose, research | ~$16,000 to $90,000 |
Kinisi's competitive advantages center on cost, speed, and reliability. The $75,000 price point (or $4,000/month rental) undercuts most bipedal humanoids. The wheeled base provides faster movement, greater stability, and lower maintenance costs compared to legged systems with their complex joint assemblies and balancing algorithms. However, the wheeled design also limits the KR1 to relatively flat, smooth surfaces, meaning it cannot climb stairs, step over obstacles, or work on uneven terrain.
Pierce has described many competitors as "show ponies," arguing that staged demonstrations in pristine laboratory environments misrepresent real-world performance. He has emphasized that Kinisi's approach focuses on deployment-ready solutions that can begin generating value for customers immediately, rather than pursuing the more ambitious but further-off goal of fully general-purpose bipedal robots.
Kinisi Robotics offers the KR1 through two commercial channels:
Direct purchase: The robot is available for approximately $75,000 USD, which positions it as one of the more affordable humanoid platforms on the market.
Monthly rental: For companies that prefer to avoid upfront capital expenditure, Kinisi offers a rental model at approximately $4,000 per month. This Robot-as-a-Service approach lowers the barrier to adoption and allows customers to evaluate the technology before committing to a purchase.
Kinisi targets its sales efforts at logistics operators, third-party logistics (3PL) providers, warehouse operators, light manufacturers, and retail fulfillment centers. The company has been recruiting partners in fulfillment, retail replenishment, and light manufacturing ahead of a planned production ramp in 2026.
As of early 2026, Kinisi Robotics is in the process of expanding from initial pilot deployments to broader commercial rollout. The company's immediate priorities include increasing the KR1's operational speed to match and eventually exceed human throughput in deployed applications, scaling production to meet demand from pilot customers, and continuing to develop the software stack that enables fleet-wide learning.
The broader market context is favorable for Kinisi's approach. The global warehouse automation market continues to grow as e-commerce expansion, labor shortages, and rising wages drive demand for robotic solutions. Kinisi's focus on affordable, practical automation for small and mid-sized warehouses positions it to serve a segment that has historically been underserved by traditional fixed automation systems.