At the micro-scale, volumetric effects such as buoyancy, gravity, and inertial forces are almost negligible. Instead, surface tension, adhesion, fluid viscosity, friction, Van der Waals forces, electrostatic and capillary forces dominate (Fearing, 1995). Van der Waals forces are always present at the micro-scale, while other forces become dominant according to the materials used to fabricate the micro-robots and the environment (Setti, 2007).

The major differences seen for the dominating effects mean that physical models developed for the control of large-scale robotics often do not apply when scaled-down to micro-robotic systems. Non-linearity brought about by the physical properties of the micro-scale world make precise, reliable, and flexible manipulation of biological micro-objects by micro-robots challenging, since micro-manipulation may be influenced by the size, geometry, and material type of the targeted objects. Furthermore, micro-robotic systems are sensitive to environmental parameters such as temperature, humidity and surface chemistry. Adding to these difficulties, in medical applications, micro-robots used for in vivo tasks must be able to handle complex environments with changing physiological conditions (e.g., soft tissue vs. blood) and unanticipated biological events. The physical models for the design of micro-robotic systems in these contexts must therefore be modified according to specific conditions under which the system will be exposed.

Beyond fundamental challenges that affect all micro-scale systems, there are also practical challenges associated specifically with the construction of micro-robotic systems. For example, reliable actuation, sensing systems and control systems are required for efficient manipulation of biological micro-objects. For micro-robots, all on-board components such as actuators, sensors, power supplies, and controllers must be miniaturized to the micron scale, which is often difficult with conventional fabrication techniques and approaches to these aspects of robot design. The exploration of new manufacturing techniques and effective control and actuation methods require further research with careful consideration of the cost and ability to handle multi-material processes for each of these areas.

To realize the full potential of micro-robots, there is a strong need for new micro-fabrication techniques, in conjunction with innovative power and energy system design, that will likely lean on novel structural design methods and advanced material development (Liu et al., 2015). Because the overall dimensions of micro-robots are too small to integrate commonly used electronic circuits, conventional strategies, such as electronic computing circuitry and software cannot be directly translated to micro-robots. New strategies are required to tackle these challenges in the assembly of advanced micro-robotics through materials, novel fabrication methods, integrated actuation and sensing capacities based on new mechanisms.

Outside of micro-robot construction, deployment of these systems for micro-manipulation applications offer yet more challenges. Due to adhesion forces, micro-objects tend to stick to micro-grippers or the substrates of the micro-robot itself. A large body of research has aimed to solve this issue by utilizing dynamical effects (Beyeler et al., 2006), gluing (Grutzeck, 2005), mechanical release (Bark et al., 2002), and gas injection (Haliyo et al., 2003). New techniques have also been developing to change the contact angle between the liquid bridge and the gripper surface via electrowetting (Vasudev and Zhe, 2008). New designs have also been proposed to reduce the contact surface between the micro-objects and the end-effectors of micro-robots by optimizing micro-gripper shape (Biganzoli et al., 2005).

Further challenges come from transferring micro-robotic systems from laboratory to clinical settings (Ongaro et al., 2016a). Laboratory-based micro-robotic techniques are increasingly mature, since most of the experiments were carried out in a homogeneous, Newtonian environment, under optical microscopes. However, for clinical practice, many other challenges need to be solved.

Micro-robots have the potential to target cancer tissues more directly and could be engineered to embed sensing and decision-making capabilities (Schmidt et al., 2020). Micro-robots for cancer targeting can broadly be classified as biologically actuated, synthetic, or hybrid if composed of cellular, synthetic, or mixed components, respectively (Gao and Wang, 2014; Thanuja et al., 2018). In all cases, active motion can be leveraged to obtain penetration and accumulation at tumour sites. Micro-robots can also be used to deliver drugs (not just in tumor tissues) in which case different mechanisms are used (e.g., chemical or enzymatic based) to control the timing and correct destination (Liu et al., 2022). This is different from in vitro applications where microscopic data can be accessed with ease. At present, most micro-robots are evaluated based on in vitro studies, while the in vivo applications are challenging.

For in vivo applications, the trackability of micro-robots with medical imaging devices become another major hurdle Aziz et al. (2020). Therefore, advanced medical imaging techniques are worth developing to track and localize micro-robots Pané et al. (2019). Biocompatibility and biodegradability have to be considered when selecting the materials to fabricate micro-robots to be used in the body Wang et al. (2018). Furthermore, how to retrieve degraded or malfunctioning micro-robots when they have conducted micro-surgical tasks is an open challenge (Ceylan et al., 2019). The development of retrievable Iacovacci et al. (2018) micro-robots is required.

For in vitro applications, micro-robots normally navigate in homogenous Newtonian fluid-like environments. However, for in vivo applications, micro-robots need to work in complex media Wu et al. (2020). For example, micro-robots commonly need to travel against blood flows, whose fluidic and physiological characteristics are non-Newtonian. Rheological properties of non-Newtonian biological fluids may limit the propulsion of micro-robots using traditional actuation techniques (Aghakhani et al., 2022). Moreover, the pH levels of the physiological environment may influence the performance of the micro-robots. Therefore, proper materials should be selected to avoid degradation in acidic environments. In addition, biological barriers, such as tight junctions and flow/rheological barriers, are known to act as obstacles to mobility (Ceylan et al., 2019). These obstaces are typically located in the entry points and may prevent micro-robots from reaching their full potential (Stillman et al.,2020).

Another challenge that micro-robots encounter during clinical translation is the host immune system and the issue of triggering innate defense mechanisms. To solve this issue, Zwitterionic materials have been used to develop non-immunogenic stealth micro-robots that avoid detection and phagocytosis from macrophages (Cabanach et al., 2020). Biohybrid micro-robots might offer good solutions to toxicity issues, while enlarging the set of possible functionalities. Structural parameters, surface chemistry and morphology of micro-robots can also be optimized to reduce the immune response by minimizing the physical interactions with the immune system (Yasa et al., 2020). In addition, micro-robots could be designed to enhance the immune system by promoting immunotherapy.

Ethical questions and risks should also be assessed before using micro-robots to act on certain organs of the human body directly. Moreover, to pass regulatory hurdles, safety and efficacy of using micro-robots must be demonstrated (Knoepfler, 2015; Agrahari et al. (2020).

Many different types of material have been assessed for the construction of micro-robotics. The materials used for the construction of micro-robotics can be categorized into i) catalytic and transient materials, ii) responsive materials, and iii) hybrid materials which combine various types of power sources.

Catalytic and transient materials include noble metals (Pt) (Lyu et al., 2021), transient metals (iron) (Karshalev et al., 2017), minerals (calcium carbonate) (Guix et al., 2016), metal alloys, semiconductors, and enzymes (Patino et al., 2018). These materials allow the gathering of chemical energy from the surrounding solutions and convert them into diverse motions. Responsive materials include magnetic materials (Zhang et al., 2019), photothermal and photocatalytic materials (Dong et al., 2018), and polarizable materials. These materials respond to the external field change or external parameters (such as temperature), which generate the mechanical force for motions (Chang et al., 2007). Hybrid materials, developed based on the combination of various power sources, take advantage of both local and external power sources as the driving force. For example, with a combination of chemical and magnetic materials, the micro-robot can convert both chemical energy and magnetic field into their driving forces (Gao et al., 2011). Here, we briefly introduce some of the most advanced and popular materials for the fabrication of micro-robotics. For example, polyvinylidene fluoride (PVDF) exhibits strong piezoelectric proprieties, wide signal bandwidth response, high pressure-voltage linearity, and high signal-to-noise ratio Lu et al. (2019), and it has been used as a promising candidate for the development of force sensors in miniaturized medical devices (Gil et al., 2020).

Emerging bio-inspired materials, for example, ionic polymer metal composite (IPMC) and electroactive polymer artificial muscle (EPAM) (Shi and Yeatman, 2021), offer great potential to produce the propulsion or bending forces required in many applications. The bending force of IPMC is generated by the effective redistribution of hydrated ions and water, which is best known as ionic-induced hydraulic actuation (Chattaraj et al., 2014). However, the current availability and implementation are hindered by the requirement for high-voltage power supplies.

Negative photoresist, such as SU-8, is an epoxy-based photo-patternable polymer with large elastic modulus, high glass transition temperature, and presents good biocompatibility Gyak et al. (2019). These properties make SU-8 a great choice for various in vivo applications, although the long-term stability in cell culture media and biocompatibility is under debate. Chronis and Wood have both developed SU-8-based micro-grippers (Wood et al., 2014), which show their promising potential in the manipulation of cells and other biological entities. Stimuli-responsive materials can naturally provide self-assembling and self-folding capabilities to micro-robots, which eliminates the need for using bulky batteries (Ongaro et al., 2016b). In this regard, improved materials property is vital to micro-robotics systems. The other intelligent materials will be introduced in Section 3.5.

For micro-robotics, all on-board components such as actuators, sensors, energy, or power units must be miniaturized to micro-scale. This requires dedicated micro-fabrication techniques.

The methods for the fabrication of microelectromechanical systems (MEMS) have enabled the rapid development of micro-robotics. MEMS allows the precise production of micro-scale features made from different materials while showing both compactness and compatibility with current industrial technologies. Surface and bulk micro-machining are among the most commonly used techniques for the fabrication of micro-robotics. For example, Volland et al. reported the micro-actuators and micro-grippers fabricated from SOI (silicon-on-insulator) wafers by using surface and bulk micro-machining (Volland et al., 2002). The advantages of this method mainly include 1) the high stability against disturbing forces and large thickness and 2) fewer fabrication steps. Lithography presents another major method that has been widely used for the fabrication of micro-robotics. As a milestone work, Kim et al. demonstrated the fabrication of three-dimensional (3D) porous structures using a 3D laser lithography system (Kim et al., 2013). After the 3D photolithography process, the micro-structures are coated with nickel and titanium films for magnetic manipulation purposes and to enhance biocompatibility. The resulting micro-robotics have been successfully demonstrated in transporting human embryonic kidney 239 (HEK 239) cells in the fluid using external magnetic fields. In addition to these two methods, laser micro-machining (Bordatchev and Nikumb, 2003), LIGA (Lithography, Electroplating, and Molding) (Menciassi et al., 2003), template electrodeposition (Li et al., 2022), thin film deposition (Hanson et al., 2022), and electro-discharge machining (EDM) (Kim et al., 2005a) have brought many concepts of micro-robots into reality.

Due to rapid technological development, 3D printing techniques have become the most promising methods for the fabrication of micro-robotics with complex shapes, helping to enhance their capabilities in conducting dexterous tasks for both in vivo and in vitro applications. 3D printing can be categorized into digital light processing (DLP) (Li and Pumera, 2021), stereolithography (SLA) (Zhan et al., 2021), fused deposition modeling (FDM) (Ben Salem et al., 2019), selective laser sintering (SLS) (Exner et al., 2008), two-photon polymerization (TPP) (Power and Yang, 2015), direct laser writing (DLW) (Tottori et al., 2012), direct metal laser sintering (DMLS) (Wong and Hernandez, 2012; Guo and Leu, 2013) and other direct writing methods (Teh, 2017). As an example, Zhang et al. developed micro-robots using the TPP technique, which can be manipulated using planar multi-spot optical tweezers. Such a versatile platform can be used for the manipulation of living cells or organisms on the microscopic scale without damage to them from direct laser illuminating. 3D printing is excellent for prototyping, but less well suited to high volume, low cost production. Therefore, advanced micro-fabrication techniques for mass production are worth developing.

Actuation is essential for micro-robots in terms of both locomotion and manipulation. Effective actuation can be achieved through the usage of advanced fabrication and the selection of proper materials. Many different types of material have been used for actuation in micro-robotics, including electrostatic (Xu, 2015), piezoelectric (Nah and Zhong, 2007), pneumatic (Alogla et al., 2015), or thermal actuators (Qu et al., 2015), and hybrid actuation systems (Chen et al., 2010b). Hybrid actuation represents the integration of different actuation modes, which combines the relative advantages of different actuation schemes for manipulating micro-robots.

Advances in miniaturized power and energy systems may also bring technological breakthroughs to the actuation of micro-robots. For example, rechargeable thin-film arbitrarily shaped batteries fabricated through semiconductor technologies can be used for micro-robots, since batteries for large-scale robots are not applicable to micro-scale robots. AbuZaiter et al. developed a novel structure of a SMA based micromanipulator for gripping purposes (Abuzaiter et al., 2016). The integration of multiple SMA bimorph actuators has been employed to manipulate objects with three degrees of freedom (DoFs). This novel design includes two links and a micro-gripper at the end of the second joint.

In addition, innovative actuation strategies have also been developed in parallel for actuating micro-robotics. For example, freeze tweezers, which are based on the Joule-Thompson throttling effect, can be used to operate micro-objects (Walle et al., 2007). Liu et al. reported a freeze tweezer using the nucleotide ice ball to manipulate micro-objects in an aqueous solution. By tuning the freezing parameters, the ice balls formed between the tweezer tip and the object can be used as contacts to pick up objects (Liu et al., 2003), where the release process during the micro-manipulation is challenging due to the sticking effect.

To tackle this challenge, non-contact manipulations have been proposed and are considered as the most promising techniques (Hattori et al., 2015). As an example of chemical controlled manipulation, Randhawa’s group presented the concept of single-use, chemically triggered, and reversible micro-grippers, which can be operated by acetic acid controlled “open” and hydrogen peroxide controlled “close” options (Randhawa et al., 2008). In addition to the chemical-controlled manipulation, the manipulation can be more precisely controlled through the finely tuned field forces. For example, magnetic fields (Chung et al., 2015), acoustic waves (Kim et al., 2015), dielectrophoresis (DEP) (Grilli and Ferraro, 2008), fluid inertial forces (Grilli and Ferraro, 2008), or microbubble streaming can be used to actuate microrobots’ microbubble (Lee et al., 2013). Zhang et al. reported the artificial bacterial flagella, which consists of a helical InGaAs/GaAs or InGaAs/GaAs/Cr tail and a thin soft-magnetic head. Such a micro-robot can swim and be manipulated in the controllable magnetic fields (Zhang et al., 2009), which represents the first generation of artificial swimmers using helical propulsion. The controlled locomotion of untethered micro-robots is attractive both for fundamental research and for biomedical applications (Sitti et al., 2015). The future development trend is to achieve fully wireless-driven manipulators for medical and biological applications under 3D control with a full six DoFs.

To avoid damage to fragile biological objects during micro-manipulation, an integrated sensing system is also required. Many micro-sensors have been designed and integrated as a part of the manipulation system. For example, Petrovic et al. developed an optical sensing system to estimate the force feedback from each mechanical micro steel arm in the micro-grippers (Petrovic et al., 2002). The integrated optical sensor includes a LED for light emission, a photosensitive element as the light receptor, and the corresponding electronics for signal processing. After calibration, the relation between the output voltage from the optical sensor and the force exerted on the tooltip of the micro-gripper can be precisely correlated. Zhu et al. reported an on-chip electrothermal sensor for high precision measurement with resolution at nanometer scale (Zhu et al., 2011).

Capacitive force sensors have been used to measure the adhesive forces between the micro-gripper fingers and the applied force (Beyeler et al., 2007). Unlike the high-resolution optical force sensor, this capacitive sensor does not require complicated setup to produce a sensitivity of 4.41 mV/μN with resolution of 70 nN. Beyeler et al. used this micro-gripper to successfully demonstrate the capacity to manipulate HeLa cells. On the other hand, piezo-resistive sensors are also capable of measuring the gripper deflection to estimate the applied forces. Molhave’s group demonstrated an electrothermally actuated micro-gripper with a capability of providing a force-feedback signal using a piezo-resistive sensor (Molhave and Hansen, 2005). This sensor is based on a simple structure consisting of three parallel beams connected by an end bar, which shows a high resolution of 0.1 mV/μN in the precise gripping of nanoscale objects. Other types of micro-sensors have been comprehensively discussed with performance metrics comparison amongst them (Wei and Xu, 2015). However, most of the micro-sensors are integrated with tethered micro-robots. As for micro-robots navigated inside human bodies, more advanced sensing systems are worth developing.

Tethered passive micro-tools are normally linked to a large-scale robotic platform for global position control or an energy delivery device. Tethered passive micro-tools such as micro-probes, micro-tweezers (Grier, 2003; Ghadiri et al., 2012), micro-tips (Xie and Régnier, 2011), and micro-pipettes (Lu et al., 2010; Bergeles and Yang, 2014) are capable of handling tiny objects (Wason et al., 2012), but rely on bulky and expensive setups (Chronis and Lee, 2005) to realize their functions without on-board actuators.

Tethered micro-robotic systems are effective tools for micro-assembly tasks, since they can ensure high precision and flexibility compared to manual operation methods. For example, a system with two tethered micro-tools has been developed to automatically grasp micro-modules to achieve bottom-up assembly of cell-laden hollow micro-tubes, which can also be used to construct other biomimetic structures with micro-structural features (Wang et al., 2016). A seven DoFs haptic interface was used to control the motions of the tethered micro-tools remotely, which can ensure safety during the micro-manipulation and enhance efficiency by providing intuitive haptic feedback to the operator (Feng et al., 2022).

More recently, a novel compliant tethered passive micro-tool has been developed for closed-loop cooperative manipulation, which was fabricated using the TPP micro-fabrication technique. It has the advantages of high flexibility, robustness, and increased workspace for both 2D and 3D object manipulation tasks. It can be used for compliant object manipulation without causing damage to delicate micro-objects (Power and Yang, 2015).

Similar to tethered passive micro-tools, tethered active micro-grippers are attached to a large-scale robotic system, but have enhanced flexibility for target object manipulation. They are typically constructed through two controllable cantilevers with on-board actuators, which can be used to grasp, clamp, or squeeze micro-objects.

As for actuation methods applied to micro-grippers, there are many different types, including electrostatic (Xu, 2015), piezoelectric (Nah and Zhong, 2007), pneumatic (Alogla et al., 2015), thermal actuators (Qu et al., 2015), SMA actuators (Munasinghe et al., 2016), or hybrid actuation systems (Chen et al., 2010b). Freeze tweezers based on the phase-changing principle can also be used (Walle et al., 2010) to operate micro-objects.

Many tethered active micro-grippers have integrated actuation and sensing capabilities (Kim et al., 2008). A novel piezoelectric actuated compliant tethered micro-gripper has been developed, which featured precise position tracking, high-speed control, and robust force regulation. This tethered active micro-gripper can be used for fast, accurate, and reliable micro-manipulation (Wang et al., 2016). Other tethered micro-grippers with integrated actuation and sensing capabilities have been reported. These include an electrostatic micro-gripper with capacitive sensor (Beyeler et al., 2006), a MEMS-based micro-gripper with electro-thermal actuators and integrated electro-thermal force sensor (Piriyanont et al., 2013), a micro-gripper with integrated electromagnetic actuator and piezoelectric force sensor (Kim et al., 2005b). The differences among those actuation and sensing strategies in terms of advantages, disadvantages, and working principles are described in another review paper with more details (Yang and Xu, 2017).

More recently, optical and multi-material fibres have provided new solutions for developing integrated devices, as they can deliver a wide range of energy levels to body tissues and have attractive features such as flexibility and cost-effectiveness in applications for microsurgery. A monolithic force-sensitive 3D micro-gripper has been designed and fabricated on the tip of a fibre using the TPP method, the dimensions of which are less than 100 μm in all three axes. It is a compliant gripper with an integrated force sensor and has a high potential for the interrogation of biological microstructures such as alveoli, villi, or even individual cells (Power et al., 2018). It also has the potential for deployment in minimally invasive microsurgical tasks or biopsy. A fluid-pressure actuated micro-gripper has been fabricated using an in situ DLW (isDLW) technique, using IP-L 780 material (Alsharhan et al., 2021). When pressure was not present, the micro-grippers remained open. The inward movement of micro-grippers was triggered by the input pressure. With different pressure inputs, the flow modes inside these microfluidic devices could be altered (Alsharhan et al., 2021).

Tethered passive micro-tools and tethered active micro-grippers are usually tethered by wires or tubes, which restrict their maneuverability when they are required to access remote environments such as inside the human body, or enclosed environments such as lab-on-a-chip devices. Therefore, untethered micro-robots are worth developing, which can make good use of wireless power transfer, or on-board power, to enable flexible locomotion or manipulation in human anatomy or confined environments.

Untethered micro-carriers can also be considered as untethered passive micro-tool, microscopic artificial swimmers (Dreyfus et al., 2005), untethered mobile micro-robots, or untethered micro-swimmers (Kim et al., 2015), whose locomotion capabilities are enabled by magnetic, acoustic, optical, or other remote physical fields for non-contact actuation (Li et al., 2017).

Light-controlled thermoplasmonic untethered micro-carriers have been developed for indirect manipulation of colloidal particles (Engay et al., 2018). Synthetic micro carriers powered by ultrasound have been reviewed in (Rao et al., 2015), while bioinspired magnetic micro-carriers have been introduced in (Bente et al., 2018). Acoustically and magnetically controlled untethered micro-carriers, consisting of micro-capsules with air bubbles trapped inside, have been fabricated by the TPP technique and coated with nickel and gold (Ren et al., 2019). They can be used for indirect single-particle manipulation, such as cell manipulation.

Untethered micro-carriers are promising for clinical translation. For example, an implantable magnetic tubular micro-robot can serve as a targeted drug delivery system in the posterior segment of the eye (Chatzipirpiridis et al., 2015b). Its mobility and controllability have been verified through experiments inside the eyes of a living rabbit (Chatzipirpiridis et al., 2015b). For targeted drug delivery, biodegradable metal-organic framework (MOF) based robots (MOFBOTs) have been developed through the combination of magnetic micro-robots and MOFs, while selective degradation in cell cultures can be achieved. To this end, flexible motility and high drug loading efficiency can be ensured for clinical applications (Terzopoulou et al., 2020).

Untethered micro-grippers should be built upon untethered micro-carriers to handle challenging tasks where manipulation is required. Different from untethered micro-carriers, untethered micro-grippers have on-broad actuators to provide gripping motions. They can perform more complex tasks with a high dexterity by making good use of integrated sensing and control systems. Untethered micro-grippers can capture and retrieve tissue or cells from hard-to-reach places that are previously infeasible. This represents an important step toward non-invasive microsurgery (Vikram Singh and Sitti, 2016).

Untethered micro-grippers require external devices for wireless power transmission to realize non-contact manipulation (Zhang and Diller, 2016; Zhang et al., 2017). An electrostatically driven untethered micro-gripper for blood vessel manipulation has been developed, which can be used for the automated study of blood vessel wall contraction forces (Wierzbicki et al., 2006). Magnetically-actuated micro-grippers have been designed for precise manipulation, which can pick and place individual microgels to target areas for assembly tasks in aquatic environments (Chung et al., 2015). These untethered micro-grippers are promising for biomedical applications such as tissue reconstruction.

Untethered micro-grippers can make diagnostic and therapeutic procedures less invasive than ever before (Nelson et al., 2010) since they can navigate through natural pathways in the human body to conduct wireless intervention (Sitti et al., 2015). For example, a hydrogel-based untethered micro-gripper actuated via a magnetic field has been developed, which is targeted for intravascular applications (Kuo et al., 2014). A thermally-driven untethered micro-gripper has also been developed (Gultepe et al., 2013a), which can navigate through narrow conduits in the human body and be used for localized biopsies. These have a high potential to be applied for complex tasks that require the cooperation of multiple micro-grippers (Zhang et al., 2017).

When it comes to power transmission for self-propelled micro-robots, one of the most favorable strategies would be to design micro-robots to harvest energy from the environment. For example, they can be powered by the chemical energy inside the cell or in the microfluidic environment. Other diverse environmental stimuli such as magnetotaxis, galvanotaxis, phototaxis, thermotaxis, and aerotaxis can be used to control the robot.

Most of the self-propulsion capabilities of micro-robots are enabled by chemical reactions. Micro-robots fabricated by catalytic materials can move quickly in an acidic environment, since the generation of continuous streams of bubbles can be used as thrust to propel and steer the micro-robot. These micro-robots can then be decomposed after task execution, and the process of retrieval is no longer necessary.

Zinc-fueled micro-robots have been fabricated, which can navigate inside the stomach of a mouse. The gastric acid reacted with a section of the micro-robot to generate non-toxic oxygen bubbles that provide propulsion (Gao et al., 2015). With the support of stimuli-responsive materials, micro-robots can generate autonomous responses in specific environments (Peraza-Hernandez et al., 2014) through self-folding (Ionov, 2013). A stimuli-responsive hybrid micro-robot fabricated by TPP has been developed (Power et al., 2017), which uses two independent actuation schemes (pH- and photo-responsive actuation systems) to control the motions of the micro-robots. It can be employed for the localization of specific cell types and target drug delivery. Related experiments have demonstrated autonomous navigation towards the target with a low-pH source in a low-flow in vitro microfluidic environment.

An untethered mobile micro-surgeon is known as the combination of a mobile micro-platform and a micro-manipulator (including end-effectors such as micro-grippers), which means that it has both manipulation and locomotion capabilities. Moreover, mobile micro-surgeons have on-board actuators, and can be self-contained and self-propelled. To this end, untethered mobile micro-surgeon can handle non-trivial tasks, and are promising to detect and combat pathologies such as cancer.

A self-folding untethered micro-robot has been developed for cell delivery (Fusco et al., 2014), which shows potential for biomedical applications such as microsurgery, on-demand drug and cell delivery. A soft thermally responsive magnetic micro-robot has been introduced in (Ongaro et al., 2016a), which can conduct pick-and-place tasks by altering both the magnetic field gradient and the temperature in the workspace of micro-robots. Biohybrid robots, developed using biomaterials, have a high potential to become mobile micro-surgeons, since they utilize power from living organisms or cells for locomotion and manipulation, while on-board actuation can be achieved due to biological activity. They have high power density and controllability (Lin et al., 2021).

Red blood cells (RBCs) have been attached to bioengineered motile bacteria to construct biohybrid robots, where RBCs were used as cargo carriers while propulsion was enabled by bacteria, with external magnetic guidance. These biohybrid micro-robots can be known as promising multimodal targeted cargo delivery systems, since they have remarkable advantages, including biocompatibility, deformability, stability. Bacteria-driven biohybrid robots have demonstrated their capabilities for active drug delivery to tumors or other target areas. More recently, Volbot (Wang et al., 2022), a type of biohybrid volvox-based micro-robot, has been developed with self-propelled capability. The biohybrid Volbot can be known as an all-in-one multifunctional micro-robot, since it has autonomous phototaxis navigation and multimode imaging capabilities. It demonstrates a high potential for applications in precision tumor therapy. More details about biohybrid micro-robots, their applications as well as the future outlook can be found in (Lin et al., 2021).

Single-cell analysis and characterization are important for new drug discovery and other relevant biomedical applications, which can be known as the killer applications for micro-robots. These applications are known as in vitro applications, since operations are conducted in laboratories outside the human body (Ahmad et al., 2021) and do not involve most of the clinical challenges as mentioned in Section 2.3. Kawahara et al. developed the quantitative evaluation of the stimulation of Pleurosira laevis and determined the relationship between the applied force and the response of a single cell (Kawahara et al., 2013). Predefined applications such as tumor-cell detection and capture tasks can be completed by untethered micro-robots that have been released into the biomedical sample (Hajba and Guttman, 2014).

Micro-robots can be used as effective tools to identify mechanical characteristics of cells, since micro-robots have advantages such as high throughput and repeatability compared to manual approaches. An untethered micro-robot equipped with a micro force sensor has been developed and used for cell positioning and sensing after force sensing calibration. The force-controlled manipulation capability of the micro-robot can ensure that the cells are positioned properly without damage or rupture (Jing et al., 2018). Magnetically driven on-chip micro-robots have been developed to investigate the response of aquatic microorganisms (Pleurosira laevis) quantitatively when providing mechanical stimulation, through which the relationship between the applied force and the response of a single cell can be analyzed (Kawahara et al., 2013).

To realize early interventional therapy, pathogens should be analyzed with rapid evaluation. Micro-robots can be effective tools to reach this target (Draz et al., 2018). Fluorescent magnetic spore-based micro-robots (FMSMs), synthesized by the direct deposition of magnetic nanoparticles, have been used as a sensing platform for toxins secreted by Clostridium difficile (C. diff). The sensing probes were encapsulated on the porous natural spores of micro-robots (Zhang et al., 2019). A biohybrid magnetic micro-robot was developed as a drug-loaded silica microtube that could be steered toward the target biofilm colony and kill E. Coli. This micro-robot consists of a magnetically guided bacteria integrated inside a drug-loaded silica microtube (Stanton et al., 2017).

Tissue biopsy coupled with the histopathologic examination is the gold standard for non-invasive diagnosis of diseases such as cancer or inflammatory diseases (Gultepe et al., 2013b), where untethered endoscopic micro-grippers can be used to retrieve tissue samples for diagnostics.

Micro-robots have been developed as effective tools for both in vitro and in vivo biopsy. For example, a set of responsive micro-grippers have been fabricated for the removal of cells from tissue embedded at the end of a capillary (Leong et al., 2009). These micro-grippers were mass-producible, while the actuation can be remotely triggered by chemicals and temperature to generate motions for in vitro biopsy, such as removing cells from tissue embedded at the end of a capillary.

More recently, untethered micro-robots are promising to conduct in vivo biopsy. In (Yim et al., 2014), a magnetically actuated untethered soft capsule endoscope has been developed, which can carry and release thermo-sensitive untethered micro-grippers (μ-grippers). These untethered micro-robots offer multi-functional strategies for gastrointestinal capsule biopsy. A stimuli-responsive soft gripper for chemomechanical controlled release was developed by Malachowski et al., which can elute chemotherapeutic drug doxorubicin while grabbing a clump of breast cancer cells (Malachowski et al., 2014).

Micro-robots can also be used for pathogen collection. A micro-robot consisting of an onion inspired multilayer structure has been developed to collect motile pathogens (Soto et al., 2020). A hollow structure serves as a trap to store the target samples, while the inner layer structure releases a chemoattractant to attract and capture motile microorganisms and store them in the hollow structure (Soto et al., 2020). However, the challenges of preserving the specimens and protecting them from contamination remain.

Due to their small size and flexibility, micro-robots are playing an increasingly important role for targeted drug delivery and cell delivery. Several typical examples of stimuli-responsive soft untethered grippers for drug delivery have been reviewed in (Ghosh et al., 2017). Dual-action micro-daggers were developed in (Srivastava et al., 2016), which can create a cellular incision followed by the drug release feature for localized drug administration. An electromagnetically actuated micro-robot has been developed for targeted drug delivery (Jeong et al., 2020), while the drug control process, including carrying, releasing, and penetrating, can be achieved by acoustically oscillating two bubbles with different natural frequency.

The in vivo study of chemically powered micro-motors has been conducted by Gao et al. (Gao et al., 2015). The body of the micro-motors can dissolve in the gastric acid gradually. The payloads carried by them were released autonomously without leaving toxicity behind. In vivo imaging and actuation of a swarm of helical micro-swimmers under rotating magnetic fields in deep tissue was developed and introduced by Servant et al. (Servant et al., 2015). The motions of the robots were tracked in real-time using fluorescence imaging. This demonstrates the possibility of optimal drug delivery to a targeted site guided by the external magnetic field (Servant et al., 2015).

As for cell delivery, magnetic field–driven micro-robots with a burr-like porous spherical structure have been introduced in (Li et al., 2018), which can serve as a promising technique for targeted therapy and tissue regeneration. The micro-robots can deliver hippocampal neural stem cells with proliferation and differentiation capabilities. To this end, in vivo cell releasing tests were conducted on mice to demonstrate its potential for cell-based therapy (Li et al., 2018).

A self-folding micro-robot, which can be used for cell delivery, has been described in (Fusco et al., 2014). Cells carried by micro-robots have been shown to reach the desired site during in vitro experiments and passing through the blood vessel-like micro-channel to arrive at the delivery site. By precisely delivering the differentiated stem cells at the pathological sites (Fox et al., 2014) or culturing human cells in 3D inside the micro-robot and using them as bio-scaffold for tissue regeneration, many diseases could be treated.

To fulfill the target of in vivo micro-surgery, untethered soft micro-grippers have significant advantages, due to the compliant motions and the capability of handling delicate objects (Rus and Tolley, 2015). They can carry cargoes and navigate in complex anatomy with varying geometries and dimensions, while being adaptive in environments where rapidly changing physiological conditions, soft tissues, and unanticipated biological events are involved.

Micro-robots can be actuated by ultrasonic power in bullet-like manner through acoustic droplet vaporization-ignition mechanisms. For example, micro-robots consisting of a hollow tube filled with perfluorocarbon emulsions can vaporize and rapidly change their state from liquid to gas to generate thrust (Kagan et al., 2012). The remarkable high speed can benefit applications, such as ablation, deep tissue penetration, etc., (Kagan et al., 2012).

An untethered magnetic micro-robot has been developed and used as an implantable tubular micro-tool for wireless ophthalmologic application (Chatzipirpiridis et al., 2015a). The micro-robot can be injected into the central vitreous humor of a living rabbit eye using a micro-needle and can be observed by an ophthalmoscope and integrated camera. More recently, micro-robots in ocular surgery have demonstrated potential applicability (Wu et al., 2018). A multi-functional biohybrid magnetite micro-robot has been developed for image-guided therapy, which has addressed the critical technical issues such as motion tracking, biocompatibility, biodegradation (Li et al., 2020), and diagnostic/therapeutic effects with preclinical in vivo experiments on subcutaneous tissue and the intraperitoneal cavity of nude mice.